This file was created by scanning the printed publication.

Errors identified by the software have been corrected;

however, some errors may remain.

United States
Department of
Agriculture Wilderness
Forest Service

Intermountain
Campsite
Forest and Range
Experiment Station Impacts:
Research Paper
INT-284
Effect of Amount
January 1982 of Use·
David N. Cole
 THE AUTHOR CONTENTS
Dr. DAVID N. COLE is a research ecologist with the Inter- Page
mountain Station's Wilderness Management research work INTRODUCTION .................................. 1
unit at the Forestry Sciences Laboratory on the University PREVIOUS STUDIES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
of Montana campus, Missoula. He is on temporary loan to STUDY AREA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
the Forest Service from the Geography Department, FIELD METHODS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
University of Oregon, Eugene. Or. Cole received his B.A. DATA ANALYSIS .................................. 9
degree in geography from the University of California, RESULTS AND DISCUSSION ...................... 10
Berkeley, in 1972. He received his Ph.D., also in geography, How Much Change Has Occurred on
from the University of Oregon in 1977. He has written the Campsites? ................................ 10
several papers on the impacts of recreational use on To What Extent Do Impacts Vary with
wilderness soils and vegetation. Differences in Amount of Use? ................. 15
How Much Change Occurs in Other Environmental
Situations? .................................... 22
Are Lakeshore Sites Particularly Fragile? ......... 25
EVALUATION OF IMPACT INDICATORS .......... 26
CONCLUSIONS AND MANAGEMENT
IMPLICATIONS ................................. 28
RESEARCH SUMMARY PUBLICATIONS CITED ............................ 31
APPENDIX: THE DISTRIBUTION OF VASCULAR
Campsites that were located near subalpine lakes in the PLANT SPECIES ENCOUNTERED ON
Eagle Cap Wilderness, Oreg., were studied. Research CAMPSITES OR CONTROL PLOTS .............. 33
objectives were to determine what ecological changes had
occurred on the sites, the extent to which amounts of
change increased with increasing use, whether lakeshore
sites had been more highly altered than sites set back from
lakeshores at least 200 ft (61 m), and the sensitivity of
selected indicators of ecological change.
Significant changes on campsites, as compared to ACKNOWLEDGMENTS
adjacent control plots included various types of
damage to mature trees; loss of seedlings; loss of under- I am grateful to many people for help with this study. Jim
growth vegetation; change in the species composition of Benedict, Sue Lindgren, Nancy Richardson, and Leslie
this undergrowth; an increase in bare mineral soil; Underhill helped in the field work. Randel Washburne
decrease in duff depth; a reduction in infiltration rates; an provided computer assistance. Charles Feddema and
increase in pH and the concentrations of magnesium, Peter Stickney identified some of the more difficult plant
calcium, and sodium ions; an increase in soil organic specimens.
matter; and an increase in soil bulk density. No difference
in the coh~entrations of potassium, phosphate, nitrate,
and total nitrogen or in soil texture could be established.
Of the 20 documented types of change, seven were more
pronounced on more heavily used sites. Of these seven,
loss of seedlings and loss of undergrowth vegetation were
almost as pronounced on light-use sites as on moderate-
or heavy-use sites, despite the statistical significance of
the relationship. The change in species composition of
the undergrowth, percent bare mineral soil, percent of
trees with exposed roots, and size of the barren campsite
core were significantly less on light-use sites than the
moderate- or heavy-use sites which had experienced
similar amounts of change. Heavy-use sites differ from
moderate-use sites primarily in the depth of the duff. This
implies that most of the change which is likely to occur on
these campsites can result from use of the site just a few
times-per-year.
Campsites set back from lakeshores had experienced
as much change as lakeshore sites. This implies that
lakeshore sites are not inherently more fragile. Where· a
lakeshore setback policy exists, other justifications for this The use of trade, firm, or corporation names in this
policy should be given. publication is for the information and convenience of the
The campsite condition class rating developed by S. S. reader. Such use does not constitute an official endorse-
Frissell proved to be the most sensitive indicator of impact ment or approval by the U.S. Department of Agriculture of
tested. Problems with this rating system, along with any product or service to the exclusion of others which may
suggested modifications, are discussed. be suitable.
United States
Department of
Agriculture Wilderness
Forest Service

Intermountain
Campsite
Forest and Range
Experiment Station Impacts:
Research Paper
INT-284
Effect of Amount
January 1982 of Use
David N. Cole

INTRODUCTION common assumption that lakeshores are more fragile than

Along with recent increases in recreational use of
wilderness has come an awareness that this use has
already modified pristine ecosystems intended for preser-
vation. In many areas, the most severe impacts occur on
campsites where use is highly concentrated, both spatially
and temporally. Managers are understandably concerned
about the,highly altered conditions of many campsites, as
it is their responsibility, according to the Wilderness Act
of 1964, to manage wilderness areas so that "natural
conditions" are preserved and "the imprint of man's work
(remains) substantially unnoticeable."
It is commonly assumed that campsite impacts are the
result of excessive use and that predicted future increases
in use will cause increasingly severe degradation. A
common response to this situation is an attempt to disperse
users from areas of concentrated use to less frequently
visited parts of the wilderness. Currently, 53 percent of all
designated wilderness units in the Forest Service and Park
Service attempt to disperse use. 1 While dispersal may
decrease campsite use and visitor encounter frequencies
in areas of heavy use, it can also increase the number of
areas where one can expect to encounter other parties,
and the number of areas which show the effects of
recreational use. This reduces the proportion of the
wilderness that offers opportunities for solitude and
shows no substantial evidence of human impact.
In order to evaluate the appropriateness of use dispersal
in various wilderness situations, or to develop any other
information-based wilderness campsite management
policy, we need a better understanding of the changes
occurring on campsites and the extent to which differences
in amounts of use affect campsite condition. A study was
designed to provide information of this kind for campsites
in Eagle Cap Wilderness in northeastern Oregon.
Permanent sampling plots were established on campsites
so that long-term changes could be evaluated. This report
describes results of the first year of study, an assessment
of changes which have already occurred, and how these
changes are related to the amount of use the site receives.
The study also compares the amount of change which
has occurred on lakeshore campsites and campsites
located more than 200 ft (61 m) from a lake. There is a
sites set back from lakes. Currently, 34 percent of all
designated wilderness units in the Forest Service and Park
Service have regulations prohibiting camping within a
certain distance of lakes. This is a sizable percentage, as
only slightly more than half of the areas in the wilderness
system contain bodies of water larger than 1 acre
(McCurdy 1977). In the case of Eagle Cap Wilderness, a
200-ft (61-m) setback has been established.
The final objective was to test the sensitivity of indicators
of impact that could be used to monitor overall site
condition. Managers have increasingly recognized the
value of monitoring systems for providing baseline
information to help them evaluate their management
programs and to identify areas where additional manage-
ment actions need to be taken. Recently, monitoring has
been mandated by Congress in the National Forest
Mangement Act. In this study, we examined the ability
of several individual measures to predict overall site
conditions and amount of change.
PREVIOUS STU Dl ES
Most detailed studies of campsite impact have been
conducted on developed campsites which are accessible
by car and which receive much heavier use than most
wilderness campsites. Backcountry campsites have been
studied in northern Minnesota (Frissell and Duncan 1965;
Merriam and others 1973), the mountains of the eastern
United States (Rechlin 1973; Bratton and others 1978),
Washington (Thornburgh 1962; Schreiner and Moorhead
1976), Idaho (Coombs 1976), and Montana (Fichtler 1980).
Of these studies, only Coombs (1976) and Fichtler (1980)
provide detailed data for a low-use area typical of most of
the wilderness in the United States.
Most studies of backcountry campsites have
documented a loss of vegetation cover and an increase
in bare ground. Changes in species composition have been
described in considerable detail (Thornburgh 1962;
1Data from a census of wilderness managers are on file at the Forestry
Sciences Laboratory. Missoula. Mont.

1
Coombs 1976; Cole 1977), as have mechanical damage to variables in which amount of impact does increase with
mature trees (Merriam and others 1973, Rechlin 1973, use, near-maximum levels of impact are usually achieved
Fichtler 1980) and the almost complete elimination of even with light use, and further increases in use do little
tree seedlings (Frissell and Duncan 1965, Coombs 1976, to aggravate the severity of these impacts (fig. 1).
Cole 1977, Fichtler 1980). Other noted changes include
an increase in soil compaction (Thornburgh 1962,
Merriam and others 1973, Cole 1977, Fichtler 1980), a
reduction in infiltration rates (Frissell and Duncan
1965), a loss of organic surface horizons (Frissell and
Duncan 1965), and erosion resulting in the exposure of
tree roots (Merriam and others 1973, Cole 1977, Fichtler
1980.)
•
In studies on developed campsites, an increase in pH
also has been a consistent finding (Young and Gilmore
1976; Dawson and others 1978; Rutherford and Scott
1979). Changes in soil nutrient concentrations have been
less consistent. Young and Gilmore (1976) found
increases in calcium (Ca), potassium (K), phosphorus (P), 1-
u
sodium (Na), and nitrogen (N), and no change in <(
magnesium (Mg) concentrations on campsites in a...
Illinois. Working in southern Ontario, Rutherford and ~
Scott (1979) found decreases in nitrate (N03), increases
in chloride (C I), and no change in phosphate (P04),
Mg, K, and sulfate (S0 4) concentrations on campsites.
Conflicting results have also been found where soil
organic matter has been studied. Dotzenko and others
(1967), Settergren and Cole (1970), Dawson and others
(1978), and Rutherford and Scott (1979) found decreases
on campsites, while Young and Gilmore (1976) and
Light Moderate Heavy
Monti and Mackintosh (1979) found increases.
In one of the few studies to relate amount of use to AMOUNT OF USE
backcountry campsite condition, Frissell and Duncan
(1965) found that more heavily used campsites in the Figure 1.--Typical research results for
Boundary Waters Canoe Area had less ground-cover the relationship between campsite
vegetation and more tree root exposure than lightly impact variables and amount of use.
used sites. They found no relationship, however, between The response of variables as use
amount of use and either vegetation loss (a measure increases from no use to light use is
based on a campsite-control comparison) or bare ground. poorly understood.
Fichtler (1980) compared impacts on lightly and heavily
used sites in Montana. He found no statistically significant
differences in amount of change in the understory, In support of this conclusion, some of the most
overstory, or soil compaction. The only significant pronounced differences between sites which receive
difference was in the amount of bare soil exposed. different amounts of use were found in the lightly used
Merriam and others (1973), working in the Boundary Idaho Primitive Area (Coombs 1976). In comparison to
Waters Canoe Area, found a poor relationship between light-use sites, heavy-use sites had considerably less
amount of use and a summary measure of campsite vegetation cover and considerably more erosion pavement.
impact. When sites were stratified by vegetation type, a This study in Eagle Cap Wilderness differentiates
more consistent relationship appeared; in each vegetation between impact on lightly, moderately, and heavily used
type, impact increased with use. The functional relation- sites which would all have been considered lightly used
ship was hyperbolic rather than linear, however, with the sites in all of the studies other than those of Coombs
rate of increase in impact decreasing as use increased. (1976) and Fichtler (1980). This focuses attention on
Similar conclusions about the nature of the relationship that part of the use spectrum which is most poorly under-
between use and impact are evident in the data presented stood (where use differences have the most pronounced
by Rechlin (1973) for backcountry campsites in the influence on amount of impact) and which is most
Adirondacks and by Dotzenko and others (1967), LaPage applicable to the wilderness situation.
(1967), Young and Gilmore (1976), Legg and Schneider
(1977), Young (1978), and James and others (1979) for
developed campsites. Although overall impact generally STUDY AREA
increases as use increases, changes in many variables,
such as infiltration rates (James and others 1979), The Eagle Cap Wilderness was selected for study
soil organic matter (Dotzenko and others 1967; Young and because it contained numerous examples of campsites
Gilmore 1976), and soil pH (Young and Gilmore 1976) are which receive light, moderate, and heavy use, in locations
not significantly correlated with amount of use. For those where at least ordinal estimates of use could be obtained.

2
Furthermore, in terms of both use and environment, the
area seemed to be representative of many heavily
glaciated, mountainous wilderness areas in the National
Forest System.
The Eagle Cap Wilderness encompasses 293,735 acres
(118 870 ha) of the Wallowa Mountains in northeastern
Oregon (fig. 2). Jagged ridges tower more than 3,280
ft (1 000 m) above deep glacial valleys which radiate from
the granitic central core of the range. Several peaks
approach 10,000 ft (3 000 m) in elevation, while the
lowest elevations in the area are under 3,600 ft (1 100m).
Over 13,000 visitors entered the Wilderness in 1978 and
accounted for about 83,000 visitor-days of use. The
distribution of use was highly concentrated, with most
visitors attracted to the more than 50 lakes scattered
through the subalpine zone (fig. 3). One area of about
2,500 acres (1 000 ha), the Lake Basin, contains 10 major
OREGON
50
50
100 150 km
100 miles
1 Figure 2.--The Eagle Cap Wilderness is in northeastern
N Oregon.
Figure 3.--Subalpine lakes are the primary destination of most visitors to the
Eagle Cap Wilderness.
3
lakes and was visited by one of every three visitors to the
Wilderness in 1978. Other lakes are more isolated and
reached by less frequently traveled trails. A few lakes
are still trailless. Twenty-two of the 26 campsites
selected for study were located at subalpine lakes where it
was possible to obtain an ordinal estimate of amount of use.
The forest overstory at all sites was dominated by Abies
lasiocarpa 2 (subalpine fir), in conjunction with Picea
engelmannii (Engelmann spruce), Pinus contorta (lodge-
pole pine), and Pinus albicaulis (whitebark pine); the
understory was usually dominated by Vaccinium
scoparium (grouse whortleberry). By confining the sample
to campsites near lakes located between 7,050 and
7,800 ft (2 150 and 2 400 m) in an Abies /asiocarpa/
Vaccinium scoparium forest type. on soils derived from
granitic bedrock, environmental differences were kept to a
minimum. Controlling environment in this manner permits
the effects of differences in amount of use to be more
precisely delineated. Four additional campsites, two in
sedge meadows above 7,800 ft (2 400 m) in elevation and
two in forests below 6,500 ft (1 981 m) in elevation, were
studied for comparative purposes.
2 All nomenclature follows Hitchcock and Conqu1st (1973)
 The amount of use each site receives had to be estimated
because no campsite-specific use data existed. This was
accomplished by assigning each lake to either a heavy-,
moderate-, or light-use category on the bases of observa-
tions and travel zone use data. The most prominenP
forested site at each lake, usually located close to where
the main trail first reaches the lake, was chosen as the site
most representative of the amount of use the lake receives.
The study sites consisted of six light-use sites, six
moderate-use sites, and 10 heavy-use sites, five within
200 ft (61 m) of a lake and five more than 200 ft (61 m)
from a lake (fig. 4-7). Sites within 200ft (61 m) of a lake
-have traditionally been the most popular and are still
frequently used, despite their having been officially closed
to camping for the last few years.
Although all analyses treated these use differences as
merely ordinal estimates, an estimate of actual amount
of use is valuable for comparative purposes. Observations
suggest that most of the light-use sites are used less
than five nights per year, with some of them receiving no
use during some years. Most moderate-use sites probably
are used 10 to 20 nights per year, while most heavy-use
sites are used 25 to 50 nights per year.

3
Prommence was defmed primarily in terms of location. We chose the site
we subjeCtively determined to be the s1te most arriving parties would
choose. We avoided automatically choosing the most heavily impacted s1te
on the iake to avoid the common circular argument 1n which heavily
impacted Sites are subjectively assigned to the heavy-use category and then
heavy-use s1tes are found to be heavily impacted

LOCATION OF SAMPLE CAMPSITES

Lake Amount of Use
1. Tombstone, moderate
2. 01 ive, Iight
3. John Henry, moderate
4. Bear, light
5. Wood, moderate
6.Chimney, heavy
7.Chimney, heavy
8. Wild Sheep, light
9. Blue, moderate
10. Hidden, light
11. Moon, I ight
12. Minam, heavy
13. Mirror, heavy
14. Horseshoe, heavy
15.Pocket, light
16. Horseshoe, heavy
17. Moccasin, heavy
18. Moccasin, heavy
19. Horseshoe, heavy
20. Glacier, moderate
21. Ice Lake Trail Jet., heavy
22. Copper Cr. Trail Jet., moderate
23. Minam, heavy
24. Cheval, light
25. Long, moderate
26. Steamboat, mode rate
0 Kilometers

0 Miles

Figure 4.--Location of sample campsites and their level of use_

4
Figure 5.--Light-use site number 8,
located at Wild Sheep Lake.

Figure 6.--Moderate-use site number 1,

located at Tombstone Lake.

Figure 7.--Heavy-use site number 7,

located at Chimney Lake.

5
FIELD METHODS
To a great extent, study methods were determined by
the need to establish permanent sampling points. This
reflected the primary goal of the study--to measure change
in campsite conditions on permanently located sites over a
5-year period. The results presented here, relating
current conditions to amount of use, were an additional
product of the study.
Each sample site consisted of both a campsite and a
similar undisturbed site in the vicinity which could serve
as a control. In each campsite, 16 linear transects were
established, radiating from a central point in 16 cardinal
directions. The distances to the edges of the campsite
and the first significant amount of vegetation were
recorded for each transect (fig. 8). The amount of
vegetation was considered significant when cover
exceeded 15 percent in a 1.09- by 3.28-ft (0.67-by 1.0(}-m)
quadrat, oriented perpendicular to and bisected by the
tape.
Within the camp area (the polygon enclosed by straight
lines connecting transect end points), all trees greater than
55 inches (140 em) tall were recorded. If damaged by
recreational use, the type of damage was recorded. Seed-
lings between 6 and 55 inches (15 and 140 em) tall were
counted within the camp polygons, exclusive of any
"islands" of undisturbed vegetation (fig. 8).
Four additional transects were established, originating
at each center point. The first transect was randomly
- - Edge of actua I camp area
----Edge of bare area
-·-Edge of camp area,
defined by transect end points
Untrampled 11 lsland 11
Figure B.--Hypothetical example of transect layout
for determining camp area and radius and bare area
and radius. Seedlings were recorded on the camp
area defined by transect end points, exclusive of
the untrampled "island;" mature trees were
recorded on the entire camp area.
6
oriented, with each subsequent transect oriented
perpendicular to the preceding one. Approximately
15 quadrats, 3.28 by 3.28 ft (1 by 1 m) were located along
these transects (fig. 9). The exact location of these
quadrats was taken from a table prepared prior to field
work, and was designed to maximize the probability that
all distances from the central point would be sampled with
equal intensity (that is, the distance between successive
quadrats on a transect decreased with distance from the
central point). This assured that (1) the entire disturbance
gradient, from central point to the undisturbed periphery,
would be equitably sampled; and (2) there was a chance
that all parts of the campsite, except the central point,
would be sampled.
In each quadrat, the coverage of ·each of the following
variables was estimated: rock, firepit, tree trunk and root,
exposed mineral soil, organic litter, and vegetation. The
cover of each vascular plant species and that of mosses as
a group were also estimated. Coverages were estimated
to the nearest percent if under 10 percent and in 10 percent
coverage classes where cover exceeded 10 percent. In
the latter case, the midpoints of each class were used to
estimate mean cover on the campsite.
Soil information was collected at four places on each
campsite between 3.28 and 6.56 ft (1 and 2 m) from the
central point (fig. 10). In contrast to the ground-cover
information, this concentrates the sampling in the most
highly disturbed parts of the campsite.
Figure 9.--Coverages of rock, firepit, tree trunk and root, exposed
mineral soil, organic litter, and vegetation were estimated in
approximately 15 quadrats on each campsite.

D Quadrat
x Soil sample location

Untrampled ''Island''

Figure 10.--Quadrat layout and location of soil

samples on hypothetical campsite.

7
 At each of the four locations, duff depth, bulk density, pH,
and infiltration rates were measured and soil samples
collected. Duff depth was a measure of the depth of the
organic litter and fermentation (0) horizons. The
colorimetric method was used to determine pH in the
field. Infiltration rates were measured with a double-
ring infiltrometer. The rate at which the first 0.39 inch
(1 em) entered the soil was called the instantaneous
infiltration rate, while the rate for the first 2 inches (5 em)
was called the saturated rate. Sample points were not
presoaked. Volumes for bulk density calculations were
determined by measuring the amount of water required to
fill a hand-excavated, cellophane-lined hole, 2 inches
(5 em) deep by about 3.5 inches (9 em) in diameter. The
excavated soil was removed in plastic bags for weight
determination (fig. 11 ). Use of the hand-excavation method
made precise volume measurements difficult, but this
method was judged to be more accurate than the use of
soil corers in these rocky soils. As with pH, bulk
density measurements were taken in the uppermost
portion of the mineral soil after the organic horizons had
been removed.
Finally, each campsite was assigned a condition class
rating, a visual estimate of campsite condition developed
by Frissell (1978).

Figure 11.--lnfiltration rates were measured with a double-ring

infiltrometer. Soil samples were collected and the volume of the
excavated hole was determined for use in calculating bulk density.

8
 Control plots were located in the vicinity in order to get
some measure of undisturbed conditions. The size of
controls varied between 980 and 2,164 fF (91 and 201 m 2 ).
Percent coverage was estimated for rock, tree trunk and
root, exposed mineral soil, organic litter, vegetation, and
plant species for the entire control plot. Seedlings were
counted on a 538-fF (5Q-m 2 ) circular subplot. Four sets
of soil measurements and samples, identical to those
taken on campsites, were taken on control plots.
DATA ANALYSIS
Distances to the edges of the campsite and the edges
of the bare area were averaged to obtain mean radius of
camp and bare area measures. They were also plotted
on scaled maps, and a polar planimeter was used to
determine the camp and bare area. "Islands" of
undisturbed vegetation intercepted by more than one
transect were subtracted from the total area.
A single mean value was calculated for duff depth, bulk
density, pH, instantaneous infiltration rate, and saturated
infiltration rate for each campsite and each control.
Infiltration rates were expressed in centimeters-per-
minute for both the 1-cm and 5-cm applications. The pH
values were converted to H+ concentrations, averaged,
and reconverted to pH values.
Soil samples were analyzed at the Montana Forest and
Conservation Experiment Station, Missoula, Mont. Each
sample was ovendried and weighed to determine bulk
density. Nitrate content was determined before drying,
using the Specific lon Analyzer. The four soil samples
from each campsite were then passed through a 2-mm
screen and composited. Calcium, potassium, magnesium,
and sodium concentrations were determined by extraction
in 1N ammonium acetate and analysis with the Atomic
Absorption Spectrophotometer. Phosphate was extracted
with dilute acid fluoride and its concentration determined
by molybdenum blue stannous chloride color reaction;
total nitrogen was determined by using the modified micro-
Kjeldahl procedure (Hesse 1972). Texture was analyzed
by buoyucous hydrometer method and organic matter
content was determined by combustion at 525° C.
For statistical analysis, standard parametric tests could
not be used because the assumption of a normally
distributed population could not be made and the sample
size was small. Therefore, nonparametric tests were used
and the tabular data presented include medians (probably
the best measure of central tendency), as well as means
and their 95 percent confidence interval.
The significance of differences between campsites and
controls was examined with the Wilcoxon matched-pairs,
signed-ranks test (Siegel 1956). Those variables that
differed were examined further in order to see if the
differences were correlated with difference in amount of
use.
The relationship between campsite impact and amount
of use can be examined in several ways. Most studies have
compared existing conditions on campsites that receive
different amounts of use. This approach has the serious
drawback of assuming that all sites were originally
identical and, therefore, that differences in the existing
conditions on I ightly and heavily used campsites reflect
9
differences in the amount of change which has occurred,
rather than differences in original conditions.
This drawback can be alleviated, to some extent, by
establishing campsite-control pairs and comparing the
differences between campsites and controls--an estimate
of amount of change--on sites which receive different
amounts of use. The problem with this approach, which
has only been tried by Frissell and Duncan (1965) and
Fichtler (1980) is that results will be misleading if control
sites are not truly similar to original campsite conditions.
Both approaches have opposing strengths and weak-
nesses. One uses no information about original conditions
(control samples), while the other uses so much informa-
tion that results may be distorted by too much faith in
control samples. Both approaches have been taken in this
study in order to profit from the unique perspective each
provides.
In this second type of analysis, both absolute and
relative measures of change are presented. Absolute
change is simply the difference between the measure on
the control site and the measure on the campsite. For
example, if vegetation cover was 10 percent on the control-
site and 1 percent on the campsite, the absolute change
would be 9 percent. Because absolute change is highly
dependent on original conditions (in the example above,
10 percent is the maximum change possible), relative
values were also calculated. Relative change is the
absolute change expressed as a percentage of the measure
on the control site. In the example above, relative change
measures show that 90 percent of the vegetation has been
lost. Again, both of these change measur:.es are pro-
vided to give as complete an interpretation of the data as
possible. Negative change values, in both cases, indicate
higher values on the campsite.
Change in species composition was measured with the
following coefficient of floristic dissimilarity:
FD = 0.5~/P1-P 2 i
where p 1 is the relative cover of a given species on the
control plot, and p 2 is the relative cover of the same species
on the campsite (Cole 1978). Additional methods of
species composition analysis are discussed in the results
section.
Finally, a summary impact rating was calculated for each
campsite in a manner similar to that employed by Merriam
and others (1973). Impact indicators included camp
area, relative vegetation loss, absolute increase in bare
ground, floristic dissimilarity, relative seedling loss,
percent of trees with trunk scars, relative decrease in duff
depth, and relative decrease in instantaneous infiltration
rate (table 1). For each of these indicators, the range in
amount of change was divided into three classes with
approximately the same number of campsites in each
class. Each campsite was assigned to one of these
classes and given an impact value of 1 to 3 (low to high
amount of change) for each indicator. The mean of all
impact values that apply to each campsite 4 determines
the index of impact or impact rating.
4
When there were no trees on the camp. the trunk scar 1nd1cator was not
considered: when there were no seedlings on controls. the seedl 1ng
1nd1cator was not considered.
 Table 1.··Values used to calculate the impact rating for each campsite 1
Indicators of impact
Relative de·
Relative Absolute Relative Relative crease in in·
Impact Camp vegetation increase in Floristic seeding Trees with decrease in stantaneous
value area loss bare ground dissimilarity loss trunk scars duff depth infiltration rate

m2 ---------------------------------------------------------------------------------------Percent---------------------------------------------------------------------------------------

<100 < 75 < 10 <50 < 76 < 15 < 45 < 20

2 100-220 75-95 10-40 50-65 76-95 15-45 45-75 20-70

3 >220 > 95 > 40 > 65 > 95 > 45 > 75 ' > 70

1
The impact rating is the mean impact value for all impact indicators used.

All correlations used Kendall's tau as a statistic (Siegel the campsite periphery than on the campsite itself. Most of
1956). Kolmogorov-Smirnov tests (Siegel1956) were used these felled trees were less than 2 inches (5 em) in diameter
to compare the amount of change which had occurred on at breast height, so the loss of saplings available to
heavily used sites located within 200 feet (61 m) of lake- eventually replace the overstory trees is also more
shores to that on heavily used sites more than 200 feet dramatic than these figures suggest.
(61 m) from lakeshores. This frequency of mechanical damage to trees is higher
than that reported elsewhere. McCool and others ( 1969)
RESULTS AND DISCUSSION report that 60 to 65 percent of the sites they studied in
the Boundary Waters Canoe Area had damaged trees; in
How Much Change Has Occurred on the Campsites?
the Eagle Cap essentially all sites with trees had damage.
Campsite area varied between 387 and 6,060 fF (36 and In two Montana backcountry areas, only 68 percent of the
563 m 2 ), with the median site being 2,077 fF (193 m 2 ) trees on all campsites had been damaged (Fichtler
(table 2). The largely barren central core comprised about 1980).
45 percent of a "typical" campsite, although this Despite this damage to the overstory, there was little
percentage varied between 15 and almost 100 percent. evidence of recreation-related tree mortality or even loss
Essentially all of the trees growing on those sites had of vigor except where trees had been felled outright. The
been "damaged" by recreationists (fig. 12). Damage to fact that more than six decades of recreational use have
many of these trees was relatively minor--lower branches had little noticeable effect suggests that premature
had been broken or nails had been driven into the trunks. mortality may never by a serious problem. Other studies
Twenty-seven percent of these trees, however, bore trunk have also noted a lack of tree mortality despite extensive
scars from chopping. Of these scars, 22 percent were mechanical damage (for example, James and others 1979)
larger than 1 fF (0.99 m 2 ) and 67 percent were located below except where the tree species is particularly susceptible
breast height, conditions under which the probability of to decay (Hinds 1976) or where severe edaphic limitations
decay in spruce and true fir is particularly high (Wright occur (Settergren and Cole 1970). This has led to the
and Isaac 1956). conclusion that mature trees are the growth form least
Another 33 percent of the trees on the campsites had sensitive to recreational impact (Leeson 1979). It is
been cut down. Observations in the field suggested that, possible, however, that premature windthrow may not
as a damage estimate, this value should have been even be recognized as recreation-caused or that disease may
higher because more felled trees were found just beyond eventually become a problem.

Table 2.··General characteristics of the campsites

Camp Bare Mutilated Trees with Felled Scarred Floristic Condition

area area trees exposed roots trees trees dissimilarity 1 class 2
Statistic (N = 22) (N = 22) (N = 19) (N = 19) (N = 19) (N = 19) (N = 22) (N = 22)

---------------------m2·-------------------- ----------------------------Percent-------------------------------------------------------------------------------------

Median 193 87 96 32 28 25 59 4.0

197 ±57 97±27 91 ± 5 34± 12 33 ± 11 27±9 55±8 3.7 ± 0.3

1
A measure of the percent change in species composition on the campsite.
2
A visual estimate of impact proposed by Frissell (1978) which varies between 1 (low impact) and 5 (high impact).
3
1ncludes a 95 percent confidence interval.

10
 Figure 12.--0ver 90 percent of the mature trees on the sample campsites
had been scarred, felled, or had limbs cut.

Of the 19 campsites with trees on the site, 17 had trees
with exposed roots. On a typical campsite, approximately
30 percent of the trees had exposed roots. Frissell
and Duncan (1965) found trees with exposed roots on
60 percent of the campsites they examined, while James
and others (1979) found 7 to 14 exposed roots per sample
tree on their campsites.
Seedling densities on control sites were almost 10-times
higher than on campsites (table 3). This amounts to a 92
percent loss of seedlings on the median campsite, with no
campsite having more than 50 percent of the seedlings
found on undisturbed sites. The few seedlings that do
survive on campsites (four seedlings per site was the
median) inevitably occur in protected areas behind

Table 3.·-Ground coverages and seedling densities for campsites and controls and estimates of amount of change 1

Bare Tree trunk

Seedlings Vegetation ground Litter Stone and root
Statistic (N = 22) (N = 22) (N =22) (N = 22) (N =22) (N =22)

Stems/ha ----------------------------------------------------------------------Percent-------------------------------------------------------------

Campsite
Median 329 6 31 59 2.4
Mean 2 507 ± 236 8±4 33 ± 10 51± 10 4.6 ± 2.4 1.5 ± 0.6

Control
Median 2,647 61 27 5.0
Mean 5,020 ± 2,362 55± 10 6±5 28±7 10.7 ± 5.4 0.8 ± 0.4

Median absolute change 2,266 47 -25 -30 3.0 -0.6

Median relative change (percent) 92 87 -1,598 -116 58 -10

3 3 3
Significance level <0.001 <0.001 <0.001 0.002 0.005

1
Absolute change is the control value minus the campsite value; relative change is the absolute change divided by the control value. Positive
values indicate that campsite values are lower than control values; negative values indicate higher campsite values than controls. Significance was tested
with the Wilcoxon matched-pairs, signed-ranks test. Differences were considered significant if the level of significance was less than 0.05. Non-
significant differences are left blank.
2
1ncludes a 95 percent confidence interval.
3
one-tailed tests were used because the direction of change was predicted prior to testing.

11
boulders or in dense clumps of saplings. There was no
evidence that any of the tree species were particularly
resistant to trampling. This near-elimination of seedlings
has been reported wherever campsite seedling densities
have been studied (Frissell and Duncan 1965; Magill and
Nord 1963; Brown and others 1977; Fichtler 1980). Along
with the loss of saplings to felling, this forecasts a future
lack of trees to replace the overstory trees when they
eventually die. Continued recruitment of trees will
probably be one of the major challenges to long-term
site maintenance.
Ground-cover changes have also been dramatic. As
vegetation cover has disappeared, increasing amounts of
bare mineral soil and organic litter have been exposed.
A small, but statistically significant, decrease in stone
cover is also evident. This may result from removal of rocks
on campsites to make fire rings or to smooth sleeping
areas, but probably also reflects a tendency for campsites
to be located on exceptionally stone-free sites.
In comparison to controls, the median campsite had one-
tenth the vegetative cover, 30 times as much bare ground,
and twice the exposed litter cover. Eighty-seven percent
of the original vegetation had been lost, leaving only 6
percent scattered about the site. These results are
comparable to those of Frissell and Duncan (1965) in the
Boundary Waters Canoe Area, and to Coombs' (1976)
measures on heavy-use sites in the Idaho Primitive Area.
Cover loss was much less extreme on developed sites in
Rhode Island (Brown and others 1977) and Pennsylvania
(LaPage 1967), where trampling-resistant species, usually
exotic grasses, maintain some vegetative cover.
The median floristic dissimilarity between campsites and
controls, 59 percent, indicates that a pronounced shift in
species composition of the undergrowth has occurred.
Inherent variability in species composition between
undisturbed stands of this vegetation type accounts for
about 25 percent of this dissimilarity (Cole 1978).
Differences in excess of 25 percent can be attributed to
changes resulting from recreational use. Previous
estimates of floristic dissimilarity on similar campsites
in the Eagle Cap Wilderness were about 80 percent (Cole
1981 a), but these campsite measurements were con-
centrated close to the heavily used central core of the
site. Apparently, and not unexpectedly, the change in
species composition on each campsite decreases from
campsite center to campsite periphery.
Every species experiences a decrease in cover on the
campsites, with the exception of three introduced species--
Poa annua (annual bluegrass), Sagina saginoides
(alpine pearlwort), and Spergularia rubra (red sand-
spurry), and six natives that are only found in small
quantities on one or two of the campsites. 5 Therefore, there
is no widespread "invasion" of species that increase in
abundance in response to recreational use. This contrasts
with lower elevation campsites where weedy invaders such
as Taraxacum officinale (common dandelion) and Poa
pratensis (Kentucky bluegrass) are usually the most
abundant species on campsites (Cole 1977).
Some species increase in relative importance on camp-
sites, however. Table 4 provides two indexes of change in
importance for the most common vascular plant species in
the area. The major species that increase in importance on
campsites are Carex microptera (small-winged sedge),
12
Carex rossii (Ross sedge), Juncus parryi (Parry's rush),
Muhlenbergia filiformis (pullup muhly), and Sibbaldia
procumbens (creeping sibbaldia), four graminoids, and a
rhizomatous, mat-forming forb with creeping stems. A
number of other studies have also found that Carex sp.,
Juncus parryi, and Sibbaldia procumbens are resistant to
trampling (for example, Landals and Scatter 1973; Holmes
and Dobson 1976).
The major species that decrease in importance on
campsites are Antennaria alpina (alpine pussytoes),
Festuca viridula (green fescue), Hieracium gracile
(slender hawkweed), Phyllodoce empetriformis (red
mountain-heath), Potentilla flabellifolia (fan-leaf
cinquefoil), Vaccinium scoparium, and Veronica cusickii
(Cusick's speedwell). The sensitivity of the brittle shrubs
Phyllodoce and Vaccinium has· been a consistent
finding (for example, Dale and Weaver 1974; Hartley
1976), but resu Its for the other species have been
inconsistent. For example, Hieracium gracille has been
judged to be resistant by Coombs (1976) and sensitive by
Schreiner (1974), Hartley (1976), and this study.
Apparently, the response of many species varies with such
factors as season, type of impact, associated plants, and,
perhaps, phenotypic variability within the species.
The three most prominent understory species on
controls-- Vaccinium scoparium, Phyllodoce
empetriformis, and Juncus parryi--undergo pronounced
shifts in importance on campsites (table 5). The median,
combined, relative cover of Vaccinium and Phyllodoce
drops from 39 percent on controls to 6 percent on camp-
sites, while the median, combined, relative cover of Juncus
and Carex rossii increases from 8 percent on controls to
28 percent on campsites. All differences are statistically
significant.
When the response of major growth forms to camping
is compared, graminoids increase in importance, while
shrubs and bryophytes decrease, and forbs are essentially
unaffected (table 5). Shrubs make up 41 percent of the
cover on controls, but only 9 percent on the median
campsite. Median graminoid values increase from 28
percent on controls to 56 percent on campsites. Similar
responses have been noted in other subalpine campsite
impact studies (Cole 1979; Leeson 1979; Weaver and others
1979).
As vegetation cover is removed, litter cover values
increase. This increase, apparent in table 3, is not a real
increase in litter; more litter is exposed because the over-
lying vegetation cover has been removed. Failure to
recognize this has caused confusion about litter response
to camping in some studies (for example, Coombs 1976).
Increased litter cover values on campsites indicate that
vegetation cover is removed more rapidly, exposing the
underlying litter, than litter is removed, exposing bare
ground. Nevertheless, litter is being removed, as the
increases in bare ground indicate. Table 3 shows that on
the median campsite, litter cover increases about 50
percent as a result of vegetation destruction, but about
25 to 30 percent of the litter cover is eroded, exposing
b3re ground.
5Appendlx 1 contains frequency and cover data for all spec1es
encountered on campsites or controls
 Table 4.··Relative importance of the most common vascular plant species 1 on campsites and controls

Species A2 82

Antennaria a/pina 0.3 0.9

Antennaria lanata 2.4 .7

Carex microptera 1.5 1.5

Carex rossii 9.0 4.5

Erigeron peregrinus .6 1.9

Festuca viridula .2 .. 9

Hieracium gracile .1 .4

Juncus parryi 3.2 1.7

Luzula hitchcockii .8 1.6

Muhlenbergia filiformis 1.5 3.6

Phyllodoce empetriformis .5 .4

Potentilla flabellifolia .5 .5

Sibbaldia procumbens 1.3 3.5

Vaccinium scoparium .2 .5

Veronica cusickii .6 .7

1
AII species with a mean cover greater than 1 percent or which occur on more than one-third of the control sites.
2
Column A is the ratio between the number of cases in which relative cover is higher on campsites than controls, and the
number of cases in which the reverse is true. Column B is the ratio between mean relative cover on campsites and mean relative
cover on controls. A number greater than 1 indicates that the species increases in relative importance on campsites.

Table 5.-·Relative cover of growth forms and selected species on campsites and controls

Growth forms Control sites Campsites

and (N =21) (N = 21) Significance
selected species Median Mean Median Mean level 1

Graminoids 28 27±7 56 55±7 0.002

Shrubs 41 42 ± 10 9 19±8 .002

Forbs 18 19±5 21 21 ±8

Bryophytes 10 12±4 5±3 .005

Carex rossii 2 2±1 8 14 ± 7 .001

Juncus parryi 6 12 ± 5 20 22±7 .005

Vaccinium scoparium 30 28±9 5 13 ± 7 .005

Phyllodoce empetriformis 9 11 ± 4 5±3 .004

1
Significance was tested with the Wilcoxon matched-pairs, signed-ranks test. Differences were considered significant if the
significance level was less than 0.05. Nonsignificant differences are left blank.

13
 This increase in bare ground is much more severe than Young and Gilmore (1976) found a 28 percent increase in
that reported by Young (1978) on developed sites in soil organic matter on forested campsites in Illinois.
Illinois, or by Frissell and Duncan (1965) and Fichtler Studies in Colorado, Iowa, and Ontario, however, have
( 1980) on backcountry sites in Minnesota and Montana, found decreases in soil organic matter on campsites
respectively. It is comparable to the results of Brown and (Dotzenko and others 1967; Dawson and others 1978;
others (1977) for developed sites in Rhode Island and of Rutherford and Scott 1979). At this time, there is no
Coombs (1976) for bare soil and erosion pavement on apparent explanation for this difference in results, as the
heavy-use, backcountry sites in Idaho. direction of change is not correlated with vegetation type,
The loss of organic litter is more evident in the decrease soil type, climatic regimen, campsite age, amount of use,
in duff depth presented in table 6. The depth of the soil or measurement technique.
organic horizons on campsites was less than one-half what Bulk density increased on campsites, but this increase
it was on controls. On the other hand, the organic matter was not as great as expected (15 percent). Bulk densities
content of the upper A horizon was about 20 percent higher were unusually low, both on campsites and controls,
on campsites than on controls, suggesting that, although reflecting the high organic matter content of the soil and
some of the surface organic litter pulverized by recreation- the influence of volcanic ash. These characteristics make
al use is probably removed by erosion, some of it moves the soil less compactible and prevent the more sizable
down into the uppermost mineral horizons where it increases of 72 percent, 46 percent, 34 percent, 30 percent,
accumulates. Monti and Mackintosh (1979) present 23 percent, and 21 percent reported on campsites in Rhode
photomicrographs which clearly show bands of "humus" Island (Brown and others 1977), Colorado (Dotzenko and
particles which have accumulated in the upper 2.54 to 7.62 others 1967), Ontario (Monti and Mackintosh 1979), Iowa
inches (1 to 3 em) of mineral soil on campsites in Ontario. (Dawson and others 1978). Missouri (Settergren and Cole
About 55 percent of the 0 horizon was lost on these camp- 1970), and Michigan (Legg and Schneider 1977), respec-
sites, a figure close to the 51 percent lost on Eagle Cap tively.
campsites, despite much heavier use in Ontario. Similar Although the effects of changes in bulk density on
measures of reduction in duff depth--60 to 65 percent-- vegetative growth are highly variable, most studies have
have been reported on campsites in the Boundary Waters shown no harmful effects until bulk densities exceed 1.3
Canoe Area (Frissell and Duncan 1965; McCool and others g/cm 3 or more (Barton and others 1966; Minore and others
1969). 1969). In fact, in some sandy soils, low levels of compaction
Measures of the magnitude and even direction of change improve the growth of certain species by increasing the
in soil organic matter content have been less consistent. water-holding capacity of the soil (Biom 1976). This
We found a 20 percent increase on campsites. Monti and suggests that, particularly on the Eagle Cap campsites
Mackintosh (1979) and Legg and Schneider (1977), where bulk densities are universally low, increases in
working in northern forest types in Ontario and Michigan, bulk density may not be a significant impact. Other
respectively, also report accumulations of organic manifestations of compaction, such as decreased infiltra-
matter in the surface mineral horizons on campsites, but tion rates or loss of microsites suitable for seed germina-
they did not compare this to conditions on control sites. tion (Harper and others 1965), may be more significant.

1
Table 6.-·Soil conditions on campsites and controls and estimates of amount of change

Ins tan· Satu·

taneous rated In·
Duff lnfiltra· filtration Total Organic Bulk
depth pH tion rate rate N0 3 K Mg Ca Na P0 4 N matter density
Statistic (N=22) (N=20) (N =20) (N =22) (N =20) (N=20) (N=20) (N=20) (N=20) (N=20) (N=20) (N=20) (N=20)

........... em/min-......... .. ....................................................... parts per million ......................................................... 3

em Percent glcm

Campsite
Median 0.25 5.48 0.33 0.16 7.6 195 61 528 53 14 3,193 18 0.95
Mean2 0.32 ± 0.18 5.44±0.16 0.38±0.10 0.16±0.04 7.9 ± 2.5 211 ±39 80±33 666 ± 222 55±5 20±9 3,493 ± 780 19±4 0.96 ±0.09

Control
Median 0.53 5.13 0.59 0.25 3.6 167 39 287 45 11 2,342 15 0.88
Mean 0.74 ± 0.30 5.13 ± 0.09 0.66 ± 0.17 0.26 ±0.05 7.4 ± 2.9 183 ± 37 42±8 3n2 ± 89 47±3 21 ±9 2,868± 626 15±2 0.88± 0.08

Median
Absolute
Change 0.30 -0.50 0.19 0.09 -4.1 -40 -23 -300 -6 0.8 -165 -2 -0.11

Median
Relative
change (percent) 51 -9 29 33 -68 -28 -107 -101 -13 -5 -20 -15

Significance 3<.001 .003 3 .003 3<.001 .001 .014 3 .047

.002 .027

1Absolute change is the control value minus the campsite value; relative change is the absolute change divided by the control value. Positive values indicate that campsite values are
lower than control values; negative values indicate higher campsite values than controls. Significance was tested with the Wilcoxon matched-pairs, signed-ranks test. Differences were con-
sidered significant if the level of significance was less than 0.05. Nonsignificant differences are left blank.
2
1nc~udes a 95 percent confidence interval.
3one-tailed test.

14
 The decreases in infiltration rates on Eagle Cap camp-
sites, while statistically significant, are also much less
pronounced than decreases found elsewhere. The rate
at which the first 0.39 inch (1 em) of water percolates into
the soil is 28 percent slower on campsites than on controls;
the decrease on campsites when 2 inches (5 em) of water
were applied was about the same--33 percent. In contrast,
other studies have found infiltration rates on controls to
be 20 to 60 times higher than on campsites (Brown and
others 1977; James and others 1979; Monti and Mackintosh
1979).
Most of this difference in magnitude of change is a
result of extremely low infiltration rates on Eagle Cap
controls. Infiltration rates on the campsites are
comparable to those on campsites in other studies; rates on
controls are an order of magnitude lower. Soils were
notably hydrophobic and were not presoaked. Hydro-
phobicity, which can cause dramatic reductions in infiltra-
tion rates, is particularly pronounced in highly organic
and sandy soils, like those in this study (Singer and
Ugolini 1976). Studies have also found soils under
ericaceous shrubs and conifers to be particularly water-
repellant (Richardson and Hole 1978).
Finally, several statistically significant changes in soil
chemistry were found. From a median value of 5.13 on
controls, pH increased to 5.48 on campsites. Concentra-
tions of Mg, Ca. and Na were also significantly higher on
campsites; Mg increased 107 percent, and Ca and Na
increased 101 percent and 13 percent, respectively.
Concentrations of N03 , K, and tot~ I N increased ~n ?~mp­
sites, but the results were so vanable that a s1gn1f1cant
difference could not be established. Phosphate content,
like particle-size distribution, 6 did not differ. These
results contrast somewhat with those of Young and
Gilmore (1976). They found similar increases inCa (116
percent) and Na (57 percent) on campsites, in addition to
increases in P (50 percent) and N (26 percent), and no
change in Mg, the nutrient which increased the most on
the Eagle Cap campsites. Rutherford and Scott (1979)
found no change in Mg, K, or P04, and a decrease in N03.
Obviously, the effects of camping on soil chemistry are
highly variable, depending upon the extent to which camp-
fire ashes are scattered about the site; the nutrient inputs
in excess food, soap, and so forth; the degree to which
leaching is reduced by decreased infiltration rates (Young
and Gilmore 1976); and the innate character of the
undisturbed soil.
It is doubtful, however, that any of these changes are
significant in an ecological sense. For example, in a
Montana study which included Abies lasiocarpa-Pinus
contorta-Carex geyeri- Vaccinium scoparium (subalpine
fir-lodgepole pine-elk sedge-grouse whortleberry) forests,
seasonal variations in soil pH and Na content were greater
than the differences between campsites and controls found
in this study (Weaver and Forcella 1979). Although
changes in Mg and Ca concentrations on campsites were
greater than their seasonal variability, the differences
were not sufficiently dramatic to suggest any basic
ecological change.
6
Data are not presented because no differences between camps1te and
control were noted
15
It is interesting to note that those ions which increased
the most on the Eagle Cap campsites are highly mobile
ions, particularly susceptible to leaching. Their presence in
high concentrations on campsites supports the hypothesis
of Young and Gilmore (1976) that campsite increases are a
result of reduced leaching.
To What Extent Do Impacts Vary with Differences
in Amount of Use?
Despite these sizable differences between campsites and
controls, the amount of change that has occurred on
campsites is extremely variable. This is reflected in the
large confidence intervals around the means in tables 2,
3, 5, and 6. It has often been assumed that most of this
variability in campsite conditions and impacts can be
attributed to differences in the amount of use the site
receives. 1 (Cole 1981 b) have suggested that environ-
mental differences usually contribute more to variability
than use differences. In this study, environmental
variability was reduced by only examining campsites in
Abies Jasiocarpa/Vaccinium scoparium forests, close to
lakes, and between elevations of 7,050 and 7,800 ft (2 150
and 2 400 m). This allowed a clearer view of the relationship
between amount of use and campsite impact.
The relationship between amount of use and impact, for
parameters which exhibit significant differences between
campsites and controls, are presented in tables 7, 8, and 9.
Amount of use is compared to existing site conditions
(table 7), the absolute amount of change which has
occurred on the site (table 8), and the relative amount of
change which has occurred on the site (table 9).
From these tables, there is no significant correlation
between the amount of use a site receives and the following
variables: camp area; mutilated, felled, or scarred trees;
litter cover; soil pH; instantaneous or saturated infiltra-
tion rates; Mg, Ca. or Na concentrations; soil organic
matter; or bulk density.
Although light-use sites are generally smaller than
moderate- and heavy-use sites (table 7), campsite area
is highly variable and there is substantial overlap in size
between use classes. For example, the low-use site at
Hidden Lake is 2,906 fF (270 m 2 ), while a heavy-use
site at Minam Lake is only 850 fF (79m 2 ). This variability
explains the nonsignificant Kendall correlation coefficient
and also the nonsignificant Kolmogorov-Smirnov contrast
between light- and moderate- use sites, despite the
difference in median and mean values. With a larger
sample size, it would probably have been possible to
conclude that light-use sites are generally smaller than
more heavily used sites (significance level was 0.064).
There is little difference in the size of moderate- and
heavy-use sites, however. Most increases in campsite
size can probably be attributed to occasional use by
abnormally large parties, or parties with packstock, a
process largely independent of frequency of use by more
typical, small, backpacking parties.
Tree mutilations also appear to increase in abundance
with increasing use, but, again, differences are dwarfed by
variability. Differences in the frequency of felled and
scarred trees are not related in any consistent manner to
the amount of use a site receives. This is not surprising
as most tree damage can probably be ascribed to a few
atypically destructive groups, again making the frequency
of use by undestructive parties irrelevant. Fichtler
(1980) also found no relationship between use and tree
InJury. James and others (1979), however, did find an
increase in the number of trunk scars per stem from 1.9
on recently built, light-use sites to 4.3 on the older, heavy-
use sites. This damage, which is cumulative, is probably
more strongly related to campsite age than to use
frequency.
Litter cover, as noted previously, is dependent on
changes in vegetation cover and bare ground. Conse-
quently, interpretation of these values is difficult.
Coombs (1976) and Fichtler (1980) reported no difference
in litter cover between light- and heavy-use sites, but they
failed to note that vegetation loss causes an apparent, but
not a real, increase in litter values. Thus, their results,
similar to these in showing no significant differences
between use levels, probably disguise a real loss in litter
cover with increasing use, coincident with an increase in
bare ground, which will be discussed. Legg and
Schneider (1977) reported a significant decrease in litter
cover on more heavily used sites, and Young (1978)
reported a reduction in litter with increased use, although
differences were not statistically significant.
Soil pH generally increases with increasing use. A large
sample size might have allowed a statistically significant
relationship to be established, but, again, differences
between use levels are minor in comparison to site-to-site
variability (table 7). The same conclusion can be drawn
from measures of the amount of change in pH on the camp-
sites (tables 8 and 9). Young and Gilmore (1976) found that
the pH on sites used 34 to 66 d'!YS per year (6.1) was
significantly higher than the pH on sites used 0 to 33 days
per year (5.7), but that a further increase in use caused
no further change. These increases, even if they can be
attributed to increased use of the campsites, are so
slight that they are not ecologically meaningful.

Table ?.··Relationship between campsite conditions and the amount of use the site receives 1

Light-use Moderate-use Heavy-use

sites sites sites
(N=6) (N=6) (N =10) Kendall
Impact parameter Median Mean Median Mean Median Mean tau

Camp area (m 2) 48 109 ± 99 224 275 ± 196 205 204±51

Bare area (m 2) 19 51± 53 122 112 ± 66 93 116 ± 40 0.30

Mutilated trees (percent) 74 60±49 85 73±40 97 93±7

Trees with exposed roots (percent) 3 7±8 33 34±26 39 39±22 .41

Felled trees (percent) 43 41 ±40 12 21 ±26 34 34± 12

Scarred trees (percent) 3 13± 17 37 36±22 11 21 ± 26

Floristic dissimilarity (percent) 31 42± 18 60 59±25 64 61 ± 11 .33

Seedlings (number/ha) 174 314 ± 366 299 404 ± 352 335 686 ± 488

Vegetation cover (percent) 9 12 ± 11 6 10± 12 4 5±2 ·.41

Bare ground (percent) 14 30±28 20 26 ± 19 35 40± 16

Litter (percent) 40 50±28 62 56±25 51 49 ± 17

Duff depth (em) 0.15 0.22 ± 0.18 0.45 0.67 ± 0.72 0.15 0.18 ± 0.09 ·.35

pH 5.25 5.32 ± 0.41 5.25 5.37 ± 0.43 5.55 5.58 ± 0.16

Instantaneous infi It ration

rate (em/min) 0.54 0.60 ± 0.19 0.19 0.28 ± 0.22 0.28 0.32 ± 0.13

Saturated infiltration
rate (em/min) 0.23 0.24 ± 0.17 0.12 0.15 ± 0.07 0.14 0.13 ± 0.04

Mg (p/m) 34 50±26 67 132 ± 237 61 77±26

Ca (p/m) 280 425 ± 265 755 1,070 ± 1,345 528 650± 263

Na (p/m) 54 52± 10 53 62± 16 51 55±8

Organic matter (percent) 12 18 ± 11 14 17 ± 13 18 20±5

Bulk density (g/cm 3 ) 0.95 1.04 ± 0.29 0.90 0.90 ± 0.19 0.95 0.94 ± 0.11

Impact rating 1.5 1.6±0.4 2.0 2.0 ± 0.4 2.2 2.1 ± 0.2 .41

1 Significance level was 0.05. Nonsignificant relationships are left blank.

16
Table a.--Relationship between the absolute amount of change which has occurred on a campsite and the amount of use the
site receives 1

Light-use Moderate-use Heavy-use

sites sites sites
(N=6) (N=6) (N =10) Kendall
Impact parameter Median Mean Median Mean Median Mean tau

Seedlings (number/ha) 1'113 1,245 ± 667 1,825 3,742±3,241 3,727 6,936 ± 5,210 0.57

Vegetation cover (percent) 37 38 ± 15 30 40±34 60 56± 15 .29

Bare ground (percent) ·5 -15 ± 26 -26 -24 ± 19 -32 -37 ± 17 -.34

Litter (percent) -26 -22±24 -26 -23± 30 -35 -25 ± 20

Duff depth (em) 0.06 0.13 ± 0.26 0.15 0.58 ± 1.21 0.35 0.49 ± 0.20 .40

pH -0.15 -0.15 ± 0.39 -0.25 -0.27 ± 0.40 -0.60 -0.36.± 0.24

Instantaneous infiltration
rate (em/min)· 0.07 0.21 ± 0.47 0.46 0.41 ±0.37 0.05 0.20± 0.23

Saturated infiltration
rate (em/min) 0.01 0.03 ± 0.12 0.12 0.16 ± 0.18 0.09 0.08 ± 0.06

Mg (p/m) -18 -19 ± 22 -28 -86 ± 136 -35 -30 ± 32

Ca (p/m) -216 -221 ± 236 ·483 -677 ± 790 -300 -206 ± 176

Na (p/m) -8 -4± 11 -15 -19± 16 -5 -6± 10

Organic matter (percent) -5 -5±6 -4 0±6 -2 -5±5

Bulk density (g/cm 3) -0.13 -0.09 ± 0.15 -0.08 -0.05±0.15 -0.14 -0.09±0.14

1
The absolute amount of change is the difference between conditions on the campsite and control. A positive change represents a decrease in that
measure on the campsite. Significance level was 0.05. Nonsignificant relationships are left blank.

Table 9.··Relationship between the relative amount of change which has occurred on a campsite and the amount of use the site receives 1

Light-use Moderate-use Heavy-use

sites sites sites
(N=6) (N=6) (N =10) Kendall
Impact parameter Median Mean Median Mean Median Mean tau

Seedlings (percent) 73 82 ± 17 92 90± 11 89 86± 11

Vegetation cover (percent) 71 78± 17 71 70±30 94 91 ±6 0.29

Bare ground (percent) -529 -504 ± 638 -1,595 -2,110±1,967 -3,293 -3,136±1,726 -.43

Litter (percent) -118 -91 ± 115 ·49 -233 ± 461 -277 -240 ± 207
,:
Duff depth (percent) 3 2± 109 21 34±52 68 72± 12 .36

pH (percent) -3 -3±8 -5 -5±8 -11 -7±5

Instantaneous infiltration
rate (percent) 8 40 ± 147 57 59±20 12 12±53

Saturated infiltration
rate (percent) -2 0±39 39 43±22 42 30±30

Mg (percent) -41 -67 ± 78 -109 -158 ± 213 -108 -78 ±50

Ca (percent) -105 -99±97 -160 -178 ± 176 -30 -69±63

Na (percent) -25 -11 ±23 -33 -45 ± 43 -10 -14±21

Organic matter (percent) -19 -43±61 -26 2±32 -20 -35 ± 35

Bulk density (percent) -16 -11 ± 14 -11 -8± 17 -16 -13 ± 16

1
The relative amount of change is the difference between campsite and control conditions expressed as a percentage of control conditions (that is,
change as a percentage of original conditions). A positive change represents a decrease in that measure on the campsite. Significance level was 0.05. Non-
significant relationships are left blank.

17
 Differences in the amount of change in other soil
properties are even more erratic and variable. For example,
instantaneous infiltration rates are reduced, on the
average, 40 percent on light-use sites, 59 percent on
moderate-use sites, and 12 percent on heavy-use sites;
standard deviations are usually greater than the differences
between use categories (table 9). The lack of relationship
between amount of use and infiltration rates, Mg
concentrations, and soil organic matter supports earlier
studies. James and others (1979) found mean infiltration
rates of 0.27 em/min on recently built, light-use sites,
and 0.29 em/min on older, heavy-use sites. As reduced
infiltration rates are probably one of the more detrimental
consequences of soil compaction, these findings--that
rates are reduced as much on light-use sites as on heavy-
use sites--seem very important.
Young and Gilmore (1976) found similar Mg concentra-
tions and organic matter content on campsites receiving
different amounts of use. Organic matter content on
their sites was 3 to 4 percent, in contrast to 17 to 20
percent on Eagle Cap campsites, suggesting that the
lack of relationship between use and organic matter
content applies to a broad range of soil types.
The lack of relationship between use and bulk density,
Ca, and Na concentrations is inconsistent with earlier
studies. Young and Gilmore (1976) report significant
differences inCa content between light- and moderate-use
sites; concentrations on light- and heavy-use sites, how-
ever, were not significantly different, and differences
between soil types were more pronounced than differences
between use categories. They also report significant
differences in Na content between light- and heavy-use
sites, but differences between light-and moderate-use
sites or moderate- and heavy-use sites were not significant.
Apparently, the relationship between soil chemistry
change and amount of recreational use can be highly
variable. The differences involved, however, appear to
almost always be so slight that differences in availability
of nutrients to plants should be negligible. This makes the
question of statistical significance moot.
The lack of relationship between amount of use and bulk
density is more difficult to dismiss, as increased bulk
density is commonly considered to be an ecologically
significant campsite impact, and both Dotzenko and
others (1967) and Legg and Schneider (1977) report
increases in bulk density associated with increased
use. Perhaps our lack of relationship is a result of measure-
ment error with the hand-excavation technique, or
perhaps the relative noncompactibility of the Eagle
Cap soils makes differences in amount of use less
important. Both Dotzenko and others (1967) and Legg
and Schneider (1977) found more pronounced differences
between controls and their light-use sites than between
light- and heavy-use sites. Increases in soil penetration
resistance on campsites, another measure of compaction,
were not correlated with amount of use in Fichtler's (1980)
study.
For the remaining parameters, there is some evidence
that impact may be related to amount of use. These
parameters will be analyzed in more detail.
SEEDLINGS
Seedling densities are actually higher on heavy-use sites
than on moderate- or light-use sites (table 7). This would
suggest that impact has been greater on light-use sites.
18
An opposing interpretation emerges when absolute
seedling loss is examined; the difference between the
density of seedlings on controls and campsites increases
from light- to heavy-use sites (table 8). Relative seedling
loss is relatively constant across the use categories.
In order to facilitate interpretation, these results have
been graphically portrayed in figure 13. Seedling densities
on controls are extremely variable, but always an order of
magnitude greater than the less variable campsite densities
(fig. 13a). The great variability in seedling densities on
control sites makes comparisons of absolute loss (fig. 13b)
misleading and favors the use of relative loss (fig. 13c)
as a measure of impact. Clearly, seedlings are almost
completely eliminated on all campsites, regardless of the
amount of use they receive. Being highly susceptible to
trampling, any consistent use is sufficient to kill most of
the seedlings. The number of seedlings surviving on a
campsite is probably more a function of the number of
protected suitable germination sites than the amount of
use the site receives. Fichtler (1980) also found no
difference in relative seedling loss between light- and
heavy-use sites.
VEGETATION COVER
In comparison to more lightly used campsites, heavy-use
sites have less vegetation (table 7), and the amount of
change, whether expressed in absolute values (table 8) or
relative values (table 9), has been greater. As shown in
figure 14, however, differences between use levels are
minor in comparison to the differences between campsites
and controls. The median cover on the light-use sites
is 9 percent; 71 percent of the original cover has been
lost. The median cover on heavy-use sites is 4 percent, a
94-percent loss. In this case, there is a statistically
significant increase in vegetation impact associated with
increased use, but the differences are not pronounced.
These results are similar to those of Frissell and Duncan
(1965), who found 12-percent vegetation cover on light-use
sites; 81 percent of the original cover had been lost. Heavy-
use sites retained a 5-percent cover, a 91 percent loss. In
their case, the difference in existing cover on campsites
was statistically significant, but the difference in amount
of change was not. Therefore, at these use levels (higher
than those found in the Eagle Cap), in the northern
Minnesota environment, differences in amount of
vegetation change were not related to amount of use.
Similarly, Fichtler (1980) found no difference in relative
cover loss between light- and heavy-use sites in Montana,
Young (1978) found no relationship between vegetation
cover and amount of use on Illinois campsites, and
LaPage (1967) found no relationship between use and
change in cover after campsites in Pennsylvania were more
than 1 year old.
The only study to show any sizable difference in cover
between light- and heavy-use sites was Coombs' (1976)
study in the Idaho Primitive Area, where use was extremely
low. She found about 30 percent cover on light-use
sites (a relative loss of 30 percent) and 8 percent on heavy-
use sites (a relative loss of 81 percent). Apparently,
vegetation change is considerably less pronounced at
very low use levels, but even a few nights of use per year
in the Eagle Cap appears to be enough to eliminate most of
the vegetation.
 (a) (a)

7500 75

<(
:::r:::
C)
z
___J
0
LL.J
5000
jt:::~an
~
0::::
LL.J
>
0
u
z
0
1-
<(
50
meant
1t median
LL.J
t/) 2500 1-
LL.J 25
C)
LL.J
>

0
C L MH CLMH
USE USE
(b) (C) (b)

<( 12, 000 90

~ ~
2::
~
0 t/)
t/)
t/)
t/)
t/)
~ t/) 0 0
___J ___J
t/) 0
t/) ___J
z z
0 8000 C) 0 0
___J
z 1- 1-
C)
;::
~
___J
z 0 LL.J
LL.J
___J
LL.J C) C)
0 t/) LL.J LL.J
LL.J
LL.J 4000 LL.J
> >
t/)
> LL.J
>
LL.J
1-
1-
:=:l
___J
::s
LL.J
1-
::s
0 0:::: LL.J
t/) 0::::
en
<(
L M H L M H L M H L M H

USE USE USE USE

Figure 13.--Seedling loss in relation to Figure 14.--Loss of vegetation cover in
amount of use. Median, mean, and two relation to amount of use. Median,
standard deviations for: (a) seedlings mean, and two standard deviations for:
per hectare on controls, light-, (a) vegetation cover on controls, light-,
moderate-, and heavy-use campsites; moderate-, and heavy-use campsites;
(b) the absolute reduction in seedling (b) absolute reduction in vegetation
density that has occurred on light-, cover on light-, moderate-, and heavy-
moderate-, and heavy-use campsites; use campsites; and (c) relative reduction
and (c) the relative reduction in seedling in vegetation cover on light-, moderate-,
density that has occurred on light-, and heavy-use campsites. The medians
moderate-, and heavy-use campsites. in each use category have been
The medians in each use category have connected by a line.
been connected by a line.

Species diversity does not vary subs~antially with
differences in amount of use. The median species rich-
ness--the number of species per 161 fF (15m 2 )-- was 12,
1 0.5, and 11.5, on light-, moderate-, and heavy-use sites,
respectively. The reciprocal of Simpson's index, a measure
of species heterogeneity which is most sensitive to
changes in dominant species (Peet 1974), was 4.76, 4.37,
and 4.48, on light-, moderate-, and heavy-use sites,
respectively. Young (1978) reported a decrease in species
richness at high-use levels, but the short-use season at
Eagle Cap precludes such heavy use. At the lower use
levels typical of high-elevation wilderness areas, species
diversity does not appear to be highly influenced by
amount of use. In fact, Coombs (1976) found more
species on light-use sites than on controls, a finding that
supports the observation that species diversity increases at
low levels of trampling stress and decreases at high levels
of stress (Slatter 1978).

19
 Change in species composition, reflected in the index
of floristic dissimilarity, does appear to increase with
increasing use, however, from a median value of 31 percent
on light-use sites to 64 percent on heavy-use sites (table
7). This suggests that, although heavy-use sites retain as
many species and almost as much cover as light-use sites,
heavy-use sites experience a more pronounced shift in
species compostion. This contrasts with the results of
Fichtler (1980), who found no relationship between floristic
dissimilarity and amount of use.
Although species composition has been more highly
altered on heavy-use campsites, there are no significant
differences in the relative importance of growth forms or
major species associated with differences in amount of
use (table 10). James and others (1979) also found no
relationship between use intensity and understory species
composition on campsites in coniferous forests, despite
pronounced shifts in species composition on campsites.
Apparently, species compositional changes may become
more pronounced as use intensity increases, but so many
other factors influence the surviving populations that the
importance of individual species and growth forms does
not vary consistently in relation to amount of use.
The more heavily used sites also have a larger central
core devoid of vegetation. This bare area increases from
a median value of 205 fF (19 m 2 ) on light-use sites to
1,313 fF (122 m 2 ) and 1,001 fF (93 m 2 ) on moderate- and
heavy-use sites, respectively (table 7). The proportion of
a site that is denuded, however, is similar on light-,
moderate-, and heavy-use sites. The barren central core is
40 percent of the area of the median light-use campsite and
45 percent of the heavy-use campsite area.
This finding contrasts with the results of Moorhead and
Schreiner (1979). Working in Olympic National Park, they
found no consistent relationship between a similar
measure of bare area and amount of use, despite significant
differences in bare area between different vegetation types.
This is one case where it has been clearly shown that
environmental differences influence the amount of impact
more highly than differences in amount of use.
BARE GROUND
Although median bare ground increases from light- to
heavy-use sites, these differences are not statistically
significant (table 7). When campsites and controls are
compared to give an estimate of amount of change,
differences are significant (tables 8 and 9). The median
light-use site has 5 percent more bare soil exposed than
its associated control (absolute change); this represents
a 529 percent increase (relative change). On moderate-use
sites, 26 percent more bare soil is exposed, a 1,595-percent
increase; and on heavy-use sites, these measures of
change increase to 32 percent· and 3,292 percent,
respectively (tables 8 and 9). Graphically portrayed in
figure 15, this suggests that the increase in bare ground
which occurs on a campsite is controlled to a significant
extent by the amount of use the site receives.
Similar results have been found by Young (1978) on
campsites in Illinois. He found that bare ground increased
from negligible amounts on controls to 30 percent on
light-use sites (0 to 33 days per year) and 56 percent on
moderate-use sites (33 to 66 days per year), but as use
exceeded 66 days per year, no significant increase in bare
ground occurred. Coombs (1976) found that bare ground
and erosion pavement increased from 2 percent on
controls to 15 percent on light-use sites and 26 percent on
heavy-use sites in the Idaho Primitive Area. Fichtler
(1980) found that bare ground increased sevenfold on
light-use sites and seventeenfold on heavy-use sites. This
was the only impact parameter that increased significantly
with increased use. These results suggest strongly that
heavy-use sites experience significantly greater increases
in bare ground exposure than light-use sites.

Table 10.--Relative cover of growth forms and selected species in relation to amount of use

Light-use Moderate-use Heavy-use

Growth forms sites sites sites
and (N =6) (N =6) (N = 10) Kendall
selected species Median Mean Median Mean Median Mean tau 1

-------------------------------------------------------------------Percent-------------------------------------------------------------------

Graminoids 66 65± 18 53 51± 15 52 51± 8

Shrubs 17 16 ± 11 9 16 ± 15 6 21 ± 14

Forbs 5 13 ± 12 35 29± 21 17 21 ± 12

Bryophytes 0.3 6± 10 3±3 2 6±5

Carex rossii 5 5±3 9 20±20 9 15±9

Juncus parryi 20 29 ± 17 12 14 ±8 27 24 ± 10

Vaccinium scoparium 3 8± 10 6 14 ± 15 6 15 ± 12

Phyllodoce empetriformis 2 6±5 0.5 1± 1 0.3 6±7

1
None of the relationships were significant at the 0.05 level.

20
 (a) (a)

60 1.5

~
0
z
=>
0
0::::
<..:)
40
rt
t. mean
median
2:
u
I
I-
0....
LI..J
0
w....
1.0
rf. t
mean
median

LLJ w....
0::::
<( 20 =>
0
0. 5
C!:l

0
C L MH C LMH
USE USE
(b) (c) (b) (C)

~ 60 ~ 4500 100
LLJ
(./) LLJ ~
(./)
~
~
(./) (./)
0:::: (./) (./)
u 0:::: 0 0
z u ___..J ___..J

40 z 3000 I I
0 1-- 1--
z 0 0.... 0....
=> z LLJ LI..J
0
0
0::::
=>
0
0
w.... w....
50
<..:) 0:::: w.... w....
LLJ
<..:) => =>
0
0::::
<(
20 LLJ
0::::
1500 0
LLJ LLJ
,:!) <( I- >
LI..J
C!:l =>
___..J 1--
I-
=>
___..J
LI..J
>
0
(./)
C!:l
::s
I.L.J
0 I- ~
(./)
ro 0 ::s
LLJ
<(

L M H
<( L M H 0:::: L M H L M H

USE USE USE USE

Figure 15.--lncrease in bare ground in rela-
tion to amount of use. Median, mean, and
two standard deviations for: (a) bare ground
cover on controls, light-, moderate-, and
heavy-use campsites; (b) absolute increase
in bare ground cover on light-, moderate-,
and heavy-use campsites; and (c) relative in-
crease in bare ground cover on light-,
moderate-, and heavy-use campsites. The
medians in each use category have been
connected by a line.
Figure 16.--Decrease in duff depth in rela-
tion to amount of use. Median, mean, and
two standard deviations for: (a) duff depth
on controls, light-, moderate-, and heavy-
use campsites; (b) absolute reduction in
duff depth on light-, moderate-, and heavy-
use campsites; and (c) relative reduction in
duff depth on light-, moderate-, and heavy-
use campsites. The medians in each use
category have been connected by a line.

DUFF DEPTH of duff increases from 3 percent on light-use sites to 21

Table 7 shows a statistically significant decrease in duff
depth associated with increased use, although light-use
sites actually have thinner organic horizons than
moderate-use sites. Measures of reduction in duff depth,
whether absolute (table 8) or relative (table 9) show sizable
increases in impact associated with increased use. All of
these measures indicate that removal of organic horizons
increases as use intensity increases, despite highly variable
campsite measurements (fig. 16). The median relative loss
percent and 68 percent on moderate- and heavy-use sites,
respectively. James and others (1979) found no relation-
ship between use intensity and duff depth, despite
pronounced reductions on campsites in general. They did
not attempt to pair campsites and controls, however;
so site-to-site variability may have disguised any relation-
ship. A second paper on the same study area (Monti and
Mackintosh 1979) showed more pronounced reductions in
duff depth on high-use sites than on medium- or low-use
sites.

21
TREE ROOT EXPOSURE
The percentage of trees with exposed roots is also
significantly lower on light-use sites--3 percent compared
to median values of 33 percent and 39 percent on
moderate- and heavy-use sites (table 7). Along with the
relatively small decrease in duff depth on light-use sites,
this suggests that loss of the surface horizons usually
does not occur until campsites are used more than about
five times per year. Conditions are highly variable,
however; some Iight-use campsites have numerous trees
with exposed roots and have lost most of their organic
horizons. James and others (1979) found that the mean
number of exposed roots within 3.28 ft (1 m) of each
sample tree increased from 7.1 on recently built, light-use
sites to 13.8 on older, heavy-use sites. Fichtler (1980),
however, found no relationship between use and linear feet
of exposed roots.
IMPACT RATING
The overall summary measure of impact, which could
vary between 1 and 3, had median values of 1.5, 2.0, and
2.2 for I ig ht-, moderate-, and heavy-use sites, respectively
(table 7). This general increase in impact in response to
increased use is statistically significant; however, there
is a considerable amount of overlap between use classes.
For example, 90 percent of the heavy-use sites have
ratings as low or lower than the rating of the most highly
impacted low-use site. In other words, heavy-use sites
are usually more highly impacted than light-use sites, but
differences are slight, and light-use sites can be more
highly impacted than heavy-use sites. Moreover,
differences between light- and moderate-use sites are
much more pronounced than differences between
moderate- and heavy-use sites.
How Much Change Occurs in Other
Environmental Situations?
It has been suggested that campsite impacts could be
reduced by locating campsites on durable sites (Cole
1981 b). Trails illustrate the importance of location where
poorly located trail segments are badly eroded, while
adjacent segments, receiving the same amount of use, are
often in good shape.
Experimental trampling studies have found that
trampling impact usually varies more between vegeta-
tion types than between use levels. For example, in a
study at Waterton Lakes National Park, 800 tramples
reduced the vegetation of a lodgepole pine stand to 1.1
22
percent of its original cover. A similar number of tramples
in a prairie grassland only reduced cover to 22.1 percent.
To maintain this percentage of original cover in the lodge-
pole stands would require a reduction in the number of
tramples to 40 (Nagy and Scotter 1974). In other words,
shifting the trampling from lodgepole to grassland sites
would accomplish as much, in terms of maintaining
original cover, as a twentyfold reduction in use of the
lodgepole sites. Moorhead and Schreiner (1979) arrived
at similar conclusions in a study of the bare area on
campsites in Olympic National Park.
In order to examine the effects of differences in
environment, campsite data are presented for four
additional sites not located close to subalpine lakes
(table 11 ). Two are located in timberline meadows close
to 8,200 ft (2 500 m) in elevation.· Campsite 15 at
Pocket Lake is a light-use site, while campsite 20 at
Glacier Lake is probably a moderate-use site. The two
other campsites are located in forested valley bottoms
below 6,500 ft (2 000 m). Campsite 21 along the West
Fork of the Wallowa River is a heavy-use site, while camp-
site 22 on the West Fork of the Lostine River is a moderate-
use site.
The two campsites in timberline meadows had lost only
10 percent of their vegetation cover, the change in
species composition had been negligible, and there was
only 3 to 4 percent more bare ground on campsites than
on controls. These measures are dramatically lower than
any of the other campsites studied. Visual impact is also
less (fig. 17), as is reflected in the low condition class
ratings of 1.0 and the impact ratings of 1.20 and 1.17.
Although there is a popular belief that timberline
meadows are unusually fragile, studies have consistently
shown the relative resistance of Carex nigricans (black
alpine sedge) meadows, the vegetation type in which
these sites are located (Campbell and Scatter 1975;
Hartley 1976). These meadows are susceptible to impact
during early summer snowmelt, but by the time they dry
enough to be usable, they are extremely durable sites.
The lower elevation campsites, in contrast, have been
as highly altered as the most heavily impacted subalpine
forest sites (fig. 18). Only two other sites have higher
overall impact ratings, and the reduction in duff depth on
these sites is particularly severe. This supports the
results of an earlier study in the Eagle Cap Wilderness
which found that more vegetation change had occurred
on forested sites, regardless of elevation, than on open
grassland or meadow sites (Cole 1981a).
Table 11.··Selected conditions and amount of change on campsites located iri alpine Carex nigricans meadows (15 and 20) and
forests below 2 000 m (21 and 22)

Campsites
Impact parameter 15 20 21 22
2
Bare area (m ) 2 101 68

Mutilated trees (percent) 100 100

Camp seedlings (number/ha) 0 0

Control seedlings (number/ha) 0 3,980

Absolute change (number/ha) 0 3,980

Relative change (percent) 0 100

Camp vegetation (percent) 88 80 3 5

Control vegetation (percent) 98 90 60 40

Absolute change (percent) 10 10 57 35

Relative change (percent) 10 11 95 88

Camp bare ground (percent) 5 4 20 35

Control bare ground (percent)

Absolute change (percent) -4 -3 -19 -34

Relative change (percent) -400 -300 -1,900 -3,400

Camp duff depth (em) 0.3 0.6 0.1

Control duff depth (em) .9 3.4 1.0

Absolute change (em) .6 2.8 .9

Relative change (percent) 67 82 90

Floristic dissimilarity (percent) 25 37 53 84

Condition class 1.0 1.0 4.0 3.5

Impact rating 1.20 1.17 2.43 2.38

23
Figure 17.--Campsite 20 is located in a timberline meadow at Glacier Lake.
Impacts were considerably less pronounced than on forested campsites.

Figure 18.--Campsite 21 is located in lower elevation forests along the West

Fork of the Wallowa River. Impacts were as pronounced as on campsites
located in subalpine forests.

24
Are Lakeshore Sites Particularly Fragile?
Lakeshore setbacks are becoming increasingly common
in wilderness. Eagle Cap Wilderness regulations prohibit
camping within 200 ft (61 m) of any lake. A common
justification for this practice is that lakeshores are more
fragile than sites set back from the lake. This supposition
was tested by comparing conditions on five heavy-use sites
located within 200 ft (61 m) of lakes and five heavy-use
sites located more than 200 ft (61 m) from lakes.
For the last few years, camping has been prohibited on
the lakeshore sites, but enforcement has been difficult. We
observed approximately the same numbers of parties still
camping on illegal sites as on legal sites. Moreover, lake-
shore sites were almost always the longest established
sites and the sites that traditionally have received the
most use. Therefore, in generalizing about the use history
of these two sets of sites, one can safely assume that
most lakeshore sites have been more heavily used for
a longer period of time, but that in recent years, both
sets have received relatively similar amounts of use.
In comparison to setback sites, lakeshore sites tend to be
somewhat larger, but less of the site is devoid of vegetation
(table 12). They have fewer seedlings, but more vegetation
cover. with a species composition that has been less highly
altered than that on setback sites. Bare ground is less
extensive, but the organic horizons are thinner. Soil pH,
Ca concentration, and bulk density are all lower on lake-
shore sites. Most of these differences are minor, however,
in comparison to highly variable site conditions.
Consequently, none of these differences were statistically
significant at the 0.05 significance level using the
Kolmogorov-Srnirnov two-sample test (Siegel 1956).

Table 12.··Campsite conditions on lakeshore sites and sites located more than 200 feet from the lakeshore

Lakeshore sites Setback sites

(N =5) (N =5)
Impact
parameter Median Mean Median Mean

Camp area (m 2 ) 219 233 ± 37 190 175 ± 76

Bare area (m 2 ) 92 108 ±32 139 123 ± 66

Seedlings (number/ha) 274 377 ± 301 637 994 ± 728

Mutilated trees (percent) 96 90 ± 11 99 95±6

Trees with exposed roots (percent) 40 38±25 38 41 ± 32

Felled trees (percent) 34 29± 13 33 38± 17

Scarred trees (percent) 17 21 ±28 5 22±24

Floristic dissimilarity (percent) 50 58± 18 66 64 ± 11

Vegetation cover (percent) 8 7±3 3 3±2

Bare ground (percent) 24 37±25 41 43 ± 15

Litter (percent) 59 50±24 50 48±20

Duff depth (em) 0.13 0.16 ± 0.08 0.20 0.20 ± 0.14

pH 5.40 5.48 ± 0.19 5.65 5.68 ± 0.13

Instantaneous infiltration rate (em/min) 0.26 0.35 ± 0.25 0.28 0.29 ±0.03

Saturated infiltration rate (em/min) 0.09 0.12 ± 0.08 0.14 0.13 ± 0.03

Mg (p/m) 63 72±29 60 81 ±39

Ca (p/m) 475 538 ± 195 587 762 ± 301

Na (p/m) 56 60 ± 13 50 49±4

Organic matter (percent) 19 23±8 17 16±4

Bulk density (g/cm 3) 0.79 0.86 ± 0.20 0.95 0.98 ± 0.07

Impact rating 2.2 2.2 ± 0.1 2.1 2.1 ±0.3

25
 When relative and absolute amounts of change were
compared, differences in the amount of change in seedling
density and pH were the only statistically significant
differences. Lakeshore sites had lost 97 percent of their
seedlings, compared to 75 percent on sites set back from
the lakes. Soil pH increased 9 percent on lakeshore
campsites, compared to 13 percent on setback sites. For
all of the other parameters measured, the amount of
change was essentially the same on the two sets of sites.
Given that seedling loss is the only impact which is more
extreme on lakeshore sites, the contention that lakeshores
are more fragile appears to be unfounded. In trampling
experiments conducted in Waterton Lakes National Park.
Nagy and Scatter (1974) found less vegetation change in
a subalpine lakeshore meadow community than in the
coniferous forests away from lakes. This is not to say,
however, that there are no justifiable reasons for pro-
hibiting camping close to lakeshores.
Although water quality studies show little evidence of
human health hazards associated with heavy use of back-
country lakes (Me Dowell 1979), there is some evidence that
ionic concentrations and benthic plant populations can be
altered by heavy use (Taylor and Erman 1979). Where lakes
are uncommon and attract abnormally large numbers of
visitors, there may be some danger that all of the lakes will
be altered by human use. In this case, the justification for
setbacks is not that lakeshores are more fragile, but that
the lake ecosystem is rare and should receive special
protection.
Another justification for setbacks is that more trails tend
to develop between campsites and the lakeshore when the
site is located close to the lake. This causes more
esthetic and ecological impact--not because the lakeshore
site is more fragile, but because the flow of traffic between
campsite and lakeshore is more destructive.
There are also a number of sociological justifications.
Lakes are commonly primary scenic attractions in
wilderness areas and should, therefore, be left as pristine
as possible. Moreover, parties camping on the lakeshore
effectively claim that territory as their own, prohibiting
other parties from having free access to the lakeshore
(Hendee and others 1977). Finally, the perception of
solitude is increased by moving people back from
lakeshores because their visibility is decreased and noise
does not carry as readily.
Prohibitions on camping close to lakes keep visitors from
camping where they most like to camp. The old, traditional
campsites were inevitably located close to lakeshores.
Managers will need good, justifiable rationales if they
are going to convince visitors to camp away from the
preferred lakeshores. The argument that mountain
lakeshores are more fragile than adjacent areas is
generally not tenable. Managers should carefully consider
other justifications for setbacks and avoid basing policy on
what is often an erroneous argument.
26
EVALUATION OF IMPACT INDICATORS
Wilderness managers have recognized a need to monitor
campsite impacts so they have some objective measure of
the changes occurring on a site. Moreover, this is now a
requirement under the National Forest Management Act.
Most monitoring programs will not be able to measure
change in the detail achieved in this study. More often,
simple, rapid techniques which can be utilized by
personnel with little training will be needed.
One simple system. developed by Frissell (1978), uses
condition classes based on visual criteria as follows:
Condition Class 1. "Ground vegetation flattened,
but not permanently inju-red.
Minimal physical change except
for.possibly a simple rock fire-
place."
Condition Class 2. "Ground vegetation worn away
around fireplace or center of
activity."
Condition Class 3. "Ground vegetation lost on most
of the site, but humus and litter
still present in all but a few
areas
Condition Class 4. "Bare mineral soil widespread.
Tree roots exposed on the
surface."
Condition Class 5. "Soil erosion obvious. Trees
reduced in vigor or dead."
We gave each campsite a condition class rating and then
correlated these ratings and other possible impact
indicators with campsite condition and change to see how
well they predicted impact (table 13).
Frissell's condition class rating was the indicator which
correlated most highly with the overall impact ratings. It
was also significantly correlated with more measures of
impact than any of the other indicators. It is not surprising
that condition class is correlated with trees with exposed
roots, vegetation cover, or bare ground because these
are characteristics used in the derivation of the rating.
The rating, however, also predicted the amount of change
in vegetation and bare ground that has occurred, as well
as the change in duff depth, floristic dissimilarity, camp
area, and bare area. These include most of the impacts
which could be noticed by visitors, as well as all of the
impacts which are related to amount of use. Soil impacts
are notably unrelated to condition class.
Camp radius and bare radius are impact indicators
originally used in Olympic National Park by Schreiner
and Moorhead (1976). Both of these indicators are
significantly correlated with the overall impact ratings,
although not as highly as condition class. Neither of
these indicators are consistently correlated with any of
the measures of impact intensity. They do, however,
provide good estimates of the areal extent of impacts.
 Table 13.--Kendall tau correlation coefficients relating campsite impact parameters and impact indicators which might potentially be utilized
in a monitoring program. Nonsignificant and redundant relationships have been left blank

Potential indicators of impact

Condition Camp Bare Vegetation Bare Trees with Impact
Impact parameter class radius radius cover ground exposed roots class

Camp area 0.38 0.50 0.35

Bare area .43 0.52 -0.37 .34

Mutilated trees .36 -.27 0.41

Trees with exposed roots .40

Floristic dissimilarity .46 .36 .26 .59

Seedlings

Seedling change (absolute) .31 .40 .24 .49

Seedling change (relative)

Vegetation cover -.48 -.35

Vegetation change (absolute) .39 .44 .37

Vegetation change (relative) .50 0.26

Bare ground .36 .32

Bare ground change (absolute) -.34 -.37

Bare ground change (relative) -.33 -.32 -.39

Duff depth .33 -.50

Duff depth change (absolute) .31 .37 .48

Duff depth change (relative) .32 .47 .54

pH

pH change (absolute) -.33 .30

pH change (relative) -.32 .32

Instantaneous infiltration -.30

Infiltration change (absolute) .35 .28

Infiltration change (relative) .35 .34

Bulk density -.36 .36

Density change (absolute) -.33 .44 -.33 -.29

Density change (relative) .39 -.33

Impact rating .50 .38 .32 .32

Vegetation cover, bare ground, and trees with exposed
roots are generally poor indicators of impact, although,
surprisingly, vegetation cover is the best indicator of soil
changes. Impact rating was included because it should
have been highly correlated with the impact measures. It
did not, however, predict amount of impact any better than
condition class, despite the time and effort required to
obtain the impact rating.
At least, for the campsites studied, it appears that
Frissell's condition class rating is the best single
indicator of campsite condition. Although it does not
identify soil changes very well, it does provide a good
indication of overall impact, as well as those changes that
use management can influence--bare ground and duff
depth, in particular. It also is correlated with the areal
extent of impacts, although a supplemental measure of
camp radius or bare radius could provide valuable
additional information.

27
 There were a number of problems with the condition
class ratings, however. The biggest problem was the
breadth of some of the categories. Despite variability
in site conditions, amount of change, and amount of use,
71 percent of the campsites received a condition class
rating of 4. In fact, much of the success of this system as
an indicator of impact may simply be its ability to separate
a few less heavily impacted sites from this majority of sites.
To be useful, Condition Class 4 will need to be subdivided
so that the concentration of consistently used campsites
in this one class is not so high.
Another major problem is that this rating does not
describe the condition of individual measurable
parameters. For example, it provides no baseline
measure of vegetation cover that could be referred to at a
later date to see if cover has changed. In other words, the
rating provides a good measure of overall campsite
condition, but little information about specific conditions.
Managers desiring more specific quantitative information
will need to use some other measure, such as percent bare
ground or vegetation cover.
Finally, many campsites could be given different ratings
depending upon the evaluative criteria chosen. For
example, some campsites had exposed tree roots
(Condition Class 4), but little bare mineral soil (Condition
Class 3). We gave these sites a rating of 3.5. Perhaps a
system of separate subjective ratings of ground vegetation,
bare mineral soil, tree root exposure, and soil erosion, the
main criteria in Frissell's system, along with a measure of
areal extent, could avoid most of these problems and still
remain highly correlated with overall impact. More
research is obviously needed.
CONCLUSIONS AND MANAGEMENT
IMPLICATIONS
The median campsite in this study has changed in the
following ways:
1. From 85 to 90 percent of the undergrowth vegeta-
tion has been removed from the 2, 150-fF (200-m 2 ) camp
area, leaving a sparse vegetation cover quite dissimilar
in composition to undisturbed sites;
2. About 30 percent of the organic litter layer has
been worn away, exposing patches of mineral soil;
3. This loss of vegetation and litter has been most
pronounced toward the center of the site where essentially
no vegetation persists and 50 percent of the organic
horizons have been removed;
4. Essentially all of the trees on the site have been
damaged, at least slightly, by recreational use; one-
fourth of the trees have been felled and another one-fourth
exhibit substantial trunk scarring;
5. Soil erosion has exposed roots on about one-third
of the trees;
6. Over 90 percent of the seedlings have been
eliminated; the few remaining seedlings are confined to
sites protected from human trampling;
7. The soil has been compacted, although even on the
most highly disturbed parts of the campsites, bulk
density has increased only 15 percent and infiltration
rates have decreased only about 30 percent; and
8. The organic content of the surface soil has
28
increased slightly, as have the pH and the concentration
of exchangeable Mg, Ca, and Na ions.
In order to utilize this information, the significance of
these changes must be questioned. In terms of an ideal
wilderness, all of these changes are significant because
they represent deviations from natural conditions. This
definition of significance cannot be practically applied to
campsites, however, because some impact is necessary
just to make the campsite functional. For example, the
shrubby understory on most of these campsites must be
removed before the site makes a comfortable sleeping area.
Given that a certain amount of impact is inevitable when-
ever a campsite receives consistent and prolonged use, a
significant impact might best be defined as any change
which reduces the future utility anq desirability of the
campsite. In other words, a significant impact would
be any change that threatens to make the site either
nonfunctional or undesirable.
Most site impacts do not appear to sharply reduce site
desirability. Although more definitive research is needed.
most evidence suggests that visitors seldom notice or are
bothered by impacts on campsites (Lucas 1979). In a
study in Yosemite National Park, for example, Lee (1975)
found that "the use of wood for fires, destruction of ground
cover, damage to trees, and other ecological changes in a
pristine environment had less influence on the visitor than
the presence of 'unnatural' objects," such as litter, horse
manure, or constructed facilities.
Of the changes found on campsites, the impact most
likely to decrease the future desirability of the campsites
is the loss of seedlings and saplings which may forecast
the eventual deforestation of campsites. A number of
studies have shown that most campers prefer campsites
that are shaded to those in the open (for example,
Cordell and James 1972). Another study has shown that,
contrary to their stated preferences, visitors to a developed
campsite usually chose largely devegetated sites (Hancock
1973). This suggests that maintenance of the overstory
is probably more important than maintenance of the under-
story. It is also more feasible. Maintenance of native
understory populations, except on protected sites, is
realistically impossible because trampling cannot be
eliminated. The overstory could be maintained by
establishing tree seedlings on protected sites, behind
logs or rocks where they will not be trampled, and then
ensuring, through an educational program, that they are
not cut down as they mature.
The impact most likely to reduce the functional ability
of a campsite is long-term erosion. While erosion was
not directly measured in this study, tree-root exposure
should provide some indication of the amount of erosion
which has occurred on a site. When these sites are
reexamined in 1984, we will have a better idea of the
magnitude of long-term erosion. Informal observations
suggest that severe erosion is rare because campsites are
usually flat and because compaction reduces the
detachability of soil particles, inhibiting erosion by
surface runoff. Managers should, however, consider
closing sites on which severe erosion is obvious.
Those sites will eventually become unusable and, at that
point, will be essentially impossible to rehabilitate.
The common assumption that deteriorated campsites
are a result of overuse is true by definition; their
deteriorated conditions are a result of "too much" use.
This study shows, however, that on the campsites
studied, even a few nights of use per year are usually
"too much," because this use causes most of the change
which is likely to occur on a campsite.
Of the 27 impact measurements taken, only seven
increase significantly in magnitude when campsite use
increases over the range of use included in this study
(that is, about a fiftyfold increase from less than one
night per year to perhaps as much as 50 nights per year).
Reductions in seedling density, vegetation cover, and duff
depth, and increases in bare ground, bare area, trees with
exposed roots, and floristic dissimilarity of the under-
growth become more pronounced as use increases. For
loss of seedlings and vegetation cover, more than 75
percent of the change occurs, however, on light-use sites
(fig. 19). In both cases, essentially any consistent annual
use eliminates almost all of the seedlings and under-
growth. Thus, the statistically significant correlation with
use does not seem to be very meaningful in either a
biological or a managerial sense.
The variables which do show meaningful differences
related to amount of use are bare ground, bare area,
duff depth, floristic dissimilarity, and trees with exposed
roots. As use increases from light to moderate amounts,
organic litter continues to be removed, creating more bare
ground and reducing duff depth; more tree roots are
exposed; the central area devoid of vegetation increases in
size; and the composition of the undergrowth continues to
change. The size of the site also appears to increase,
although this difference was not statistically significant.
As use surpasses moderate amounts--probably 10 to 20
nights per year--further increases in use cause little
additional change in any of the variables other than duff
depth, which continues to decrease dramatically with
increased use.
The two major implications of these results are:
1. Any annually repeated use of campsites in this
environmental situation will cause major onsite ecological
changes;
2. Even fiftyfold use reductions will do relatively little to
reduce campsite impacts.
These results only apply strictly to campsites near lake-
shores, in Abies lasiocarpa/Vaccinium scoparium forests
of the Eagle Cap Wilderness, between 7,050 and 7,800 ft
(2 150 and 2 400 m), on soils derived from granite bed-
rock. Similar results, however, were also reported by
Fichtler (1980) in two Montana areas. Working in Abies
lasiocarpa forests, with an undergrowth differenUrom that
in Eagle Cap on volcanic soils between 5,600 and 7,250
ft (1 700 and 2 200 m), he found tf.lat the only significant
difference between light- and heavy-use sites was more
exposed mineral soil on heavy-use sites.
Less fragile sites, such as low-elevation grasslands,
could probably support more use before near-maximum
levels of impact were achieved, but even on a low-
elevation campground in Pennsylvania in "abandoned
field" vegetation on deep, well-drained and productive
silt loam flood plain soils, LaPage (1967) found that "the
relationship between barren ground and cumulative man-
days of use weakened and disappeared entirely" after the
campground was 3 years old. More research is needed, but
it appears clear that low levels of annual use are sufficient
to cause most of the change which is likely to occur on a
site. In relatively fragile high-elevation forests where a
large proportion of wilderness campsites are located, this
threshold appears to be no higher than a few nights of use
per year.
In deciding how best to manage impacts, it is important
to distinguish between impact intensity and the total
aggregate area of impact. The conclusion of this study
is that increasing or reducing use has very Iitie effect

(a) (b) (c) (d)

100
~ ~
Vl
0-
Vl
Vl
Vl
0 z£ £
Figure 19.--The relationship between a- 0 -1 =>LJ...J Iz
-1
z OVl 1-o
a.._
mount of use and amount of impact for 0
z 50 ~ ~~ LJ...JI-
ou
those variables with a statistically signifi- 1- LJ..JO::
w... =>
.....J
0 ;::: e:::u
<(Z
w...o
:::lLJ..J
cant relationship: (a) percent reduction in LJ...J
LJ...J
LJ...J
0 co - Cl 0::
Vl LJ...J
seedling density; (b) percent reduction in 0 >
vegetative cover; (c) absolute increase in
L M H L M H L M H L M H
bare ground; (d) percent reduction in duff
depth; (e) bare area; (f) percent of trees with (e) (f) (g)
exposed roots; and (g) floristic dissimilarity.
100
For each use category, the median change Cl
LJ...J ~
is expressed as a percentage of the highest ~ Vl >-
o- u::
median value for any use category. ~ a..£ -a:::
0::
<( 50 ~~ ~5
LJ...J :s:o
0 0::~
0::
~0::
o_
<(
co LJ..J ~Vl
0:: Vl
1- Cl
0
L M H L M H L M H

USE USE USE

29
on impact intensity, such as the magnitude of vegetation
loss. Impact intensity can be more effectively minimized
through improved campsite location and changes in
visitor behavior.
The good condition of campsites in Carex nigricans
meadows, in relation to forested sites, suggests the
value of improved campsite location. Schreiner and
Moorhead (1976), for example, have shown that the bare
area of a campsite varies primarily with differences in
vegetation. Many studies have shown that sites dominated
by graminoids will usually suffer less vegetation loss than
sites dominated by shrubs, such as those in this study.
Managers might consider closing badly deteriorated
sites, particularly those experiencing severe erosion, and
opening new sites in more durable locations if necessary.
Visitor behavior can be changed through either
minimum-impact camping education or regulations on
type of use. A change in visitor behavior could eliminate
the scarring and felling of trees and reduce some of the
impact on soii chemistry through more careful use of fire
and reduced pollution. On the other hand, it can do little
to reduce vegetation and litter loss, the increase in bare
ground, and soil compaction. These are inevitable
consequences of use that are probably acceptable to
most visitors, and, in most cases, pose no threat to the
long-term usefulness or desirability of the campsite. It
might be worth informing people of the problem with tree
reproduction, however, in the hope that they might be
careful to avoid trampling established seedlings.
Regulations on type of use could also be useful. If
effective, a campfire prohibition could at least reduce
damage to live trees and changes in soil chemistry. Keep-
ing horses out of camp areas could reduce damage to trees,
trampling of seedlings and other undergrowth, soil
erosion and tree root exposure, and the size of campsites.
Managers must decide if the potential for improvement in
site conditions is worth the imposition of regulations.
In contrast to its limited effect on the intensity of
impacts on existing campsites, use redistribution could
have a pronounced effect on the total aggregate area of
impacts. In areas where the amount of use is high enough
that annually repeated use of campsites--even once a
year in many places--is likely to occur, the area of impact
could be reduced by encouraging repetitive use of fewer
campsites. Use dispersal, in this situation, will usually
increase the number of deteriorated campsites,
with little compensatory improvement in conditions on
former heavy-use sites. In the Eagle Cap Wilderness,
for example, we heard complaints that many areas that
were pristine a few years ago now have impacted camp-
sites.
In areas that receive at least moderate amounts of use,
such as all the lakes in this study reached by trail,
30
repetitive use of few sites could probably be achieved by
merely asking people to camp on previously used sites.
In most areas, it should not be necessary to officially
designate "legal" campsites, as many areas in the National
Park Service and Parks Canada do. Impacts could then be
further reduced by closing some campsites in areas having
more campsites than necessary. Some lakes in the Eagle
Cap Wilderness, for example, are encircled by more than
100 campsites.
In areas where use is very low, dispersal could minimize
impacts. This policy requires extreme caution, however,
to insure that repeat use of sites does not occur before the
site can recover. If it does, campsites will deteriorate over
time and the number and total aggregate area of campsite
impacts will increase greatly.
A program of education in minimum-impact camping
techniques is a prerequisite for a dispersal policy.
Visitors must be well educated before dispersal will work.
Many National Park Service and Parks Canada areas also
prohibit fires in areas of dispersed camping to reduce the
potential for campsite change. A final necessity is a
monitoring program capable of evaluating how well the
program is working.
Many types of monitoring programs could be suggested.
In highly dispersed-use settings, an inventory of sites
showing signs of human use may be adequate. In areas
which receive more consistent use, more information on
site conditions would be desirable. If a manager does not
need quantitative baseline data on specific campsite
conditions, such as amount of vegetation cover, some
modified 7 version of Frissell's condition classes would
provide a good measure of overall site condition. When
combined with a measure of campsite or bare area, an
inventory of campsites utilizing these two measures
should enable the manager to identify trends both in the
intensity of impact on individual sites and the areal spread
of impacts, either through the enlargement of sites or the
pro I iteration of new campsites.
In many cases, however, managers may need more
detailed information on site conditions. Measurements of
this type are much more costly because they require
precise replication of previous measurements. Our
research suggests that percent bare ground measurements
are probably the most valuable because, unlike most
impact parameters, bare ground varies in response to
amount of use. Managers could attempt to manage use in
such a manner that bare ground does not increase. Once
the cost of precise replication is accepted, however,
additional measurements are relatively cheap and
should be given serious consideration. In particular,
some measure of the campsite or bare area should be taken
to supplement measures of impact intensity.
7 1n addit1on to modifying some of the overly broad categor1es.
managers may also need to redef1ne categones to more accurately reflect
the1r environmental s1tuat1on. As suggested previously. a system of separate
subJeCtive rat1ngs of ground vegetat1on. bare mmerai soli. tree root
exposure. and soli eros1on. the ma1n cr1ter1a of Fr1ssell s system. along w1th
any addit1onal parameters of concern. m1ght be particularly useful
PUBLICATIONS CITED
Barton, Howard, Wayne G. McCully, Howard M. Taylor,
and James E. Box, Jr.
1966. Influence of soil compaction on emergence and
first-year growth of seeded grasses. J. Range Manage.
19:118-121 .
Blom, C. W. P. M.
1976. Effects of trampling and soil compaction on the
occurrence of some Plantago species in coastal sand
dunes. Oecol. Plant. 11:225-241.
Bratton, Susan Power, Matthew G. Hickler, and James H.
Graves.
1978. Visitor impact on backcountry campsites in the
Great Smoky Mountains. Environ. Manage. 2:431-442.
Brown, J. H., Jr., S. P. Kalisz, and W. R. Wright.
1977. Effects of recreational use on forested sites.
Environ. Geol. 1:425-431.
Campbell, S. E., and G. W. Scatter.
1975. Subalpine revegetation and disturbance studies,
Mt. Revelstoke National Park. Can. Wild!. Serv. Rep.,
Edmonton, Alta. 99 p.
Cole, David Naylor.
1977. Man's impact on wilderness vegetation: an
example from Eagle Cap Wilderness, northeastern
Oregon. Ph.D. diss. Univ. Oregon, Eugene. 307 p.
Cole, David N.
1978. Estimating the susceptibility of wildland vegeta-
tion to trailside alteration. J. Appl. Ecol. 15:281-286.
Cole, David N.
1979. Reducing the impact of hikers on vegetation: an
application of analytical research methods. In
Recreational impact on wildlands, conference pro-
ceedings. p. 71-78. R. Ittner and others, eds. USDA
For. Serv., Pac. Northwest Reg., R-6-001-1979.
Cole, David N.
1981 a. Vegetational changes associated with
recreational use and fire suppression in the Eagle Cap
Wilderness: some management implications. Bioi.
Conserv. 20:247-270
Cole, David N.
1981 b. Managing ecological impacts at wilderness
campsites: an evaluation of techniques. J. For. 79:
86-89.
Coombs, Elizabeth A. K.
1976. The impact of camping on vegetation in the
Bighorn Crags, Idaho Primitive Area. M.S. thesis.
Univ. Idaho, Moscow. 63 p.
Cordell, Harold K., and George A. James.
1972. Visitors' preferences for certain physical
characteristics of developed campsites. USDA For.
Serv. Res. Pap. SE-1 00, 21 p. Southeast. For. Exp.
Stn., Asheville, N.C.
Dale, D., and T. Weaver.
1974. Trampling effects on vegetation of the trail
corridors of north Rocky Mountain forests. J. Appl.
Ecol. 11:767-772.
Dawson, J. 0., D. W. Countryman, and R. R. Fittin.
1978. Soil and vegetative patterns in northeastern Iowa
campgrounds. J. Soil and Water Conserv. 33:39-41.
Dotzenko, A. D., N. T. Papamichos, and D. S. Romine.
1967. Effects of recreational use on soil and moisture
conditions in Rocky Mountain National Park. J. Soil
and Water Conserv. 22:196-197.
31
Fichtler, Richard K.
1980. The relationship of recreational impacts on
backcountry campsites to selected Montana habitat
types. M.S. thesis. Univ. Montana, Missoula. 109 p.
Frissell, Sidney S.
1978. Judging recreation impacts on wilderness camp-
sites. J. For. 76:481-483.
Frissell, Sidney S., Jr., and Donald P. Duncan.
1965. Campsite preference and deterioration in the
Quetico-Superior canoe country. J. For. 63:256-260.
Hancock, "H" K. ,
1973. Recreation preferences: its relation to user
behavior. J. For. 71:336-337 .•
Harper, J. L., J. T. Williams, and G. R. Sagar.
1965. The behavior of seeds in soi I. I. The heterogeneity
of soil surfaces and its role in determining the
establishment of plants from seed. J. Ecol. 53:273-286.
Hartley, Ernest Albert.
1976. Man's effects on the stability of alpine and sub-
alpine vegetation in Glacier National Park, Montana.
Ph.D. diss. Duke Univ., Durham, N.C. 258 p.
Hendee, John C., Roger N. Clark, and Thomas E. Dailey.
1977. Fishing and other recreation behavior at high
mountain lakes in Washington state. USDA For. Serv.
Res. Note PNW-304, 27 p. Pac. Northwest For. and
Range Exp. Stn., Portland, Oreg.
Hermann, Frederick J.
1970. Manual of the Carices of the Rocky Mountains
and Colorado Basin. USDA For. Serv. Agric. Handb.
374, 397 p. Washington, D.C.
Hesse, P. R.
1972. A textbook of soil chemical analysis. 520 p.
Chemical Publ. Co., New York.
Hinds, Thomas E.
1976. Aspen mortality in Rocky Mountain campgrounds.
USDA For. Serv. Res. Pap. RM-164, 20 p. Rocky Mt.
For. and Range Exp. Stn., Fort Collins, Colo.
Hitchcock, C. Leo, and Arthur Cronquist.
1973. Flora of the Pacific Northwest. 730 p. Univ.
Washington Press, Seattle.
Holmes, Daniel 0., and Heidi E. M. Dobson.
1976. Ecological carrying capacity research: Yosemite
National Park. Part 1. The effects of human trampling
and urine on subalpine vegetation, a survey of past
and present backcountry use, and the ecological
carrying capacity of wilderness. U.S. Dep. Commerce,
Natl. Tech. lnf. Cent. PB-270 955, 247 p.
James, T. D. W., D. W. Smith, E. E. Mackintosh, M. K.
Hoffmann, and P. Monti.
1979. Effects of camping recreation on soil, jack pine
(Pinus banksiana), and understory vegetation in a
northwestern Ontario park. For. Sci. 25:233-249.
Landals, M., and G. W. Scatter.
1973. Visitor impact on meadows near Lake O'Hara,
Yoho National Park. Can. Wild I. Serv. Rep., Edmonton,
Alta. 184 p.
LaPage, Wilbur F.
1967. Some observations on campground trampling and
ground cover response. USDA For. Serv. Res. Pap.
NE-68, 11 p. Northeast. For. Exp. Stn., Broomall, Pa.
Magill, Arthur W., and Eamor C. Nord.
1963. An evaluation of campground conditions and
needs for research. USDA For. Serv. Res. Note PSW-4,
8 p. Pac. Southwest For. and Range Fxp. Stn.,
Berkeley, Calif.
Merriam, L. C., Jr., C. K. Smith, D. E. Miller, and others.
1973. Newly developed campsites in the Boundary
Waters Canoe Area--a study of five years' use. Univ.
Minn., St. Paul, Agric. Exp. Stn. Bull. 511, 27 p.
Minore, Don, Clark E. Smith, and Robert F. Woollard.
1969. Effects of high soil density on seedling root
growth of seven northwestern tree species. USDA For.
Serv. Res. Note PNW-112, 6 p. Pac. Northwest For. and
Range Exp. Stn., Portland, Oreg.
Monti, P., and E. E Mackintosh.
1979. Effect of camping on surface soil properties in
the boreal forest region of northwestern Ontario.
Soil Sci. Soc. Am. J. 43:1024-1029.
Moorhead, Bruce B., and Edward S. Schreiner.
1979. Management studies of human impact at back-
country campsites in Olympic National Park,
Washington. In Proc., Cont. on scientific research
in the National Parks. p. 1273-1278. R. M. Linn, ed.
U.S. Dep. Interior, Natl. Park Serv. Trans. Proc. Serv.
5. Washington, D. C.
Nagy, J. A. S., and G. W. Scatter.
197 4. A quantitative assessment of the effects of human
and horse trampling on natural areas, Waterton
Lakes National Park. Can. Wild!. Serv. Rep.,
Edmonton, Alta. 145 p.
Peet, Robert K.
1974. The measurement of species diversity. Annu.
Rev. Ecol. and Syst. 5:285-308.
Rechlin, Michael A.
1973. Recreational impact in the Adirondack high
peaks wilderness. M.S. thesis. Univ. Mich., Ann Arbor.
65 p.
Richardson, J. L., and F. D. Hole.
1978. Influence of vegetation on water repellency in
selected western Wisconsin soils. Soil Sci. Soc. Am.
J. 42:465-467.
Rutherford, G. K., and D. C. Scott.
1979. The impact of recreational land use on chemistry
in a provincial park. Park News 15:22-25.
32
Schreiner, Edward George.
1974. Vegetation dynamics and human trampling in
three subalpine communities of Olympic National
Park, Washington. M.S. thesis. Univ. Wash., Seattle.
150 p.
Schreiner, E. S., and B. B. Moorhead.
1976. Human impact studies in Olympic National Park.
In Proc. symp. on terrestrial and aquatic ecological
studies of the Northwest. p. 59-66. East. Wash. State
Coli., Cheney.
Settergren, C. D., and D. M. Cole.
1970. Recreation effects on soil and vegetation in the
Missouri Ozarks. J. For. 68:231-233.
Siegel, Sidney.
1956. Nonparametric statistics for the behavioral
sciences. 312 p. McGraw-Hiii:New York.
Singer, M. J., and F. C. Ugolini.
1976. Hydrophobicity in the soils of Findley Lake,
Washington. For. Sci. 22:54-58.
Slatter, R. J.
1978. Ecological effects of trampling on sand dune
vegetation. J. Bioi. Educ. 12:89-96.
Taylor, T. P., and E. C. Erman.
1979. The response of benthic plants to past levels of
human use in high mountain lakes in Kings Canyon
National Park, California, USA. J. Environ. Manage.
9:271-278.
Thornburgh, Dale Alden.
1962. An ecological study of the effects of man's
recreational use at two subalpine sites in western
Washington. M.S. thesis. Univ. Calif. Berkeley. 50 p.
Weaver, Tad, Donn Dale, and E. Hartley.
1979. The relationship of trail condition to use,
vegetation, user, slope, season, and time. In Recre-
ational impact on wildlands, conference proceedings.
p. 94-101. R. Ittner and others, eds. USDA For. Serv.,
Pac. Northwest Reg., R-6-001-1979.
Weaver, T., and F. Forcella.
1979. Seasonal variation in soil nutrients under six Rocky
Mountain vegetation types. Soil Sci. Soc. Am. J.
43:589-593.
 APPENDIX

THE DISTRIBUTION OF VASCULAR PLANT SPECIES ENCOUNTERED ON CAMPSITES OR CONTROL PLOTS.

2
Frequency of occurrence 1 Mean cover
2
Mean relative cover
Species Camps Controls Camps Controls Camps Controls
Achillea millefolium 0 3 0 + 0 0.4
Agrostis thurberiana 0 + 0 0 +
Agrostis variabilis 4 4 + .4 0.4 .3
Allium validum 0 2 0 + 0 +
Antennaria alpina 4 16 0.2 1.0 1.5 1.6
Antennaria lanata 9 14 .3 2.1 2.5 3.4
Antennaria microphyl/a 0 1 0 + 0 .1
Arabis lyallii 1 0 + 0 .7 d
Arabis sp. 0 1 0 + 0 .1
Arenaria aculeata 0 5 0 .1 0 .4
Arnica cordifo/ia 3 0 .1 0 .3
Arnica mol/is 1 5 .1 .5 .1 .5
Arnica parryi 0 1 0 + 0 +
Aster alpigenus 5 4 .1 .3 1.4 .4
Carex geyeri 1 2 .1 .1 .8 .2
Carex luzulina 0 1 0 + 0 +
Carex microptera 4 5 .2 1. 7 2.6 1.7
Carex nigricans 2 + .5 .1 .8
Carex rossii 19 20 .6 1.2 13.4 3.0
Carex scopulorum 0 2 0 0.5 0 0.5
Carex spectabilis 3 5 6 0.3 .3 2.6 .7
Cassiope mertensiana 2 3 + .3 + .2
Castilleja chrysantha 1 6 + .2 + .2
Danthonia intermedia 3 6 + .4 .4 .4
Deschampsia caespitosa 0 2 0 .6 0 .5
Dodecatheon alpinum + + + +
Epilobium alpinum 2 2 + + + +
Epilobium angustifolium 1 6 + .1 + .1
Epilobium sp. 0 0 + 0 +
Erigeron peregrinus 9 16 .1 1.2 2.9 1.5
Eriogonum flavum piperi 0 0 + 0 .3
Eriogonum ovalifolium 3 + + .3 +
Festuca viridula 4 14 .3 1.6 2.1 2.4
Gaultheria humitusa 1 5 + .6 .1 .6
Gayophytum humile 2 + + + +
Gentiana calycosa 0 2 0 .2 0 +
Hieracium albertinum 2 0 + 0 .3
Hieracium gracile 4 14 + .3 .3 .6
Holodiscus discolor 0 1 0 + 0 .1
Hypericum anagalloides 0 + 0 + 0
Hypericum formosum 0 6 0 0.4 0 0.4
Juncus drummondii 1 3 + + .2 +
Juncus mertensianus 0 0 + 0 +
Juncus parryi 20 21 1.7 6.2 21.2 12.4
Ledum glandulosum 4 + .7 + .6
Lewisia pygmaea 0 + 0 + 0
Ligusticum tenuifolium 4 5 + .1 + .1
Linanthastrum nuttallii 0 0 + 0 +
Lonicera utahensis 0 0 + 0 +
Luzula campestris 0 1 0 + 0 +
Luzula hitchcockii 7 9 .4 1.8 4.7 2.9
Muhlenbergia filiformis 9 10 .2 .9 2.5 .7

Continued

33
 APPENDIX (Con.)

THE DISTRIBUTION OF VASCULAR PLANT SPECIES ENCOUNTERED ON CAMPSITES OR CONTROL PLOTS.

2 2
Frequency of occurrence 1 Mean cover Mean relative cover
Species Camps Controls Camps Controls Camps Controls
Oryzopsis exigua 1 6 + .1 .8 .5
Osmorhiza chilensis 0 0 + 0 +
Parnassia fimbriata 0 0 + 0 +
Pedicularis contorta 0 1 0 + 0 +
Penstemon fruticosus 0 2 0 + 0 .2
Penstemon rydbergii 1 4 0 .3 .4 .6
Phleum alpinum 4 4 + .3 .•2 .2
Phyl/odoce empetriformis 12 18 .3 8.8 4.5 10.6
Poa annua 1 0 + 0 .1 0
Poa gracillima 0 0 + 0 +
Poa leibergii 0 1 0 + 0 .1
Poa sandbergii 0 1 0 + 0 +
Poa sp. 0 2 0 + 0 +
Polemonium pulcherrimum 4 + 0.1 + 0.4
Polygonum phytolaccaefolium 4 + .2 1.2 .4
Potentilla diversifolia 0 1 0 + 0 +
Potentilla flabellifolia 5 6 .1 2.1 .9 1.8
Potentilla glandulosa 0 0 + 0 .1
Potentilla gracilis glabrata 0 1 0 + 0 +
Ranunculus eschscholtzii 0 2 0 + 0 +
Ranunculus populago 1 1 + + + +
Sagina saginoides 2 0 .1 0 .9 0
Senecio cymbalarioides 0 2 0 .5 0 .3
Spergularia rubra 1 0 + 0 + 0
Trisetum spicatum 0 6 0 .4 0 .7
Trisetum wolfii 0 0 .2 0 .1
Vaccinium caespitosum 4 .1 .7 .5 1.5
Vaccinium scoparium 19 21 .6 19.1 12.6 26.4
Veratrum viride 1 2 + + .2 +
Veronica cusickii 9 14 .2 1.9 1.9 2.6
Veronica serpyllifolia 1 0 + 0 + 0
Veronica wormskjoldii 1 0 + 0 + 0
Viola adunca 3 5 + .3 .1 .3

1Number of sites out of a maximum of 22 on which the species was found.

2Mean cover is the actual canopy coverage of the species, while relative cover expresses actual cover as a percentage of the total
cover on the site. A plus ( +) indicates less than 0.1 percent cover.

3This species was determined by Charles Feddema to be C. tolmei Boott, which Hitchcock and Cronquist (1973) consider to be
synonymous with C. spectabilis. Some authorities equate C. tolmei with C. paysonis Clokey (Hermann 1970).

'- U.S. GOVERNMENT PRINTING OFFICE: 1982-576-026/140 Region No.8

34
Cole, David N.
1982. Wilderness campsite impacts: effect of amount of use. USDA For.
Serv. Res. Pap. INT-284, 34 p. lntermt. For. and Range Exp. Stn., Ogden,
Utah 84401.

Subalpine lakeshore campsites were studied in the Eagle Cap Wilderness,

Oreg. Light-use campsites had experienced almost as much alteration as
moderate- and heavy-use sites. Sites set back from lakeshores had changed
as much as lakeshore sites. Selected indicators of ecological change were
evaluated. Implications of this research to management of wilderness
campsites are discussed.

KEYWORDS: ecological impact, campsites, wilderness management

 The Intermountain Station, headquartered in Ogden,
Utah, is one of eight regional experiment stations charged
with providing scientific knowledge to help resource
managers meet human needs and protect forest and range
ecosystems.
The Intermountain Station includes the States of
Montana, Idaho, Utah, Nevada, and western Wyoming.
About 231 million acres, or 85 percent, of the land area in the
Station territory are classified as forest and rangeland. These
lands include grasslands, deserts, shrublands, alpine areas,
and well-stocked forests. They supply fiber for forest in-
dustries; minerals for energy and industrial development; and
water for domestic and industrial consumption. They also
provide recreation opportunities for millions of visitors each
year.
Field programs and research work units of the Station
are maintained in:

Boise, Idaho
Bozeman, Montana (in cooperation with Montana
State University)
Logan, Utah (in cooperation with Utah State
University)
Missoula, Montana (in cooperation with the
University of Montana)
Moscow, Idaho (in cooperation with the U niver-
sity of Idaho)
Provo, Utah (in cooperation with Brigham Young
University)
Reno, Nevada (in cooperation with the University
of Nevada)